STATUS OF CORAL REEFS IN THE WESTERN ATLANTIC: RESULTS OF INITIAL SURVEYS, ATLANTIC AND GULF RAPID REEF ASSESSMENT (AGRRA) PROGRAM Edited by Judith C. Lang ATOLL RESEARCH BULLETIN Issued by NATIONAL MUSUEM OF NATURAL HISTORY SMITHSONIAN INSTITUTION WASHINGTON, D.G., U.S.A. JULY 2003 Atlantic and Gulf Rapid Reef Assessment (AGRRA) A Joint Program of The Rosenstiel School of Marine and Atmospheric Science, University of Miami and The Ocean Research and Education Foundation Inc. Organizing Committee Co-chairpersons: Robert N. Ginsburg & Philip A. Kramer Patricia Richards Kramer Judith C. Lang Peter F. Sale Robert S. Steneck Database Manager, Kenneth W. Marks Regional Advisors Pedro M. Alcolado, Cuba Claude Bouchon Guadaloupe Jorge Cortes, Costa Rica Janet Gibson, Belize J. Ernesto Arias-Gonzalez, Mexico Zelinda Margarita Leao, Brazil Citations of papers in this volume: Feingold, J.S., S.L. Thornton, K.W. Banks, N.J. Gasman, D. Gilliam, P. Fletcher, and C. Avila. 2003. A rapid assessment of coral reefs near Flopetown, Abaco Islands, Bahamas (stony corals and algae). Pp. 58-75 in J.C. Lang (ed.). Status of Coral Reefs in the western Atlantic: Results of initial Surveys, Atlantic and Gulf Rapid Reef Assessment (AGRRA) Program. Atoll Research Bulletin 496. Cover Design: Hunter Augustus Cover Photograph: Martin Moe ATOLL RESEARCH BULLETIN NO. 496 STATUS OE CORAL REEFS IN THE WESTERN ATLANTIC: RESULTS OF INITIAL SURVEYS, ATLANTIC AND GULF RAPID REEF ASSESSMENT (AGRRA) PROGRAM EDITED BY JUDITH C. LANG SUPPORT FOR THE PUBLICATION OF THIS VOLUME PROVIDED BY THE BACARDI FAMILY FOUNDATION, NATIONAL CENTER FOR CARIBBEAN CORAL REEF RESEARCH AND OCEAN RESEARCH AND EDUCATION ISSUED BY NATIONAL MUSUEM OF NATURAL HISTORY SMITHSONIAN INSTITUTION WASHINGTON, D.C., U.S.A. JULY 2003 11 ACKNOWLEDGMENT The Atoll Research Bulletin is issued by the Smithsonian Institution to provide an outlet for information on the biota of tropical islands and reefs and on the environment that supports the biota. The Bulletin is supported by the National Museum of Natural History and is produced by the Smithsonian Press. This issue was financed by the Bacardi Family Foundation, the National Center for Caribbean Coral Reef Research (NCORE) , and the Ocean Research and Education Foundation (ORE) . Distribution was accomplished with funds from Atoll Research Bulletin readers and authors. The Bulletin was founded in 1951 and the first 117 numbers were issued by the Pacific Science Board, National Academy of Sciences, with financial support from the Office of Naval Research. Its pages were devoted largely to reports resulting from the Pacific Science Board's Coral Atoll Program. All statements made in papers published in the Atoll Research Bulletin are the sole responsibility of the authors and do not necessarily represent the views of the Smithsonian nor of the editors of the Bulletin. Articles submitted for publication in the Atoll Research Bulletin should be original papers in a format similar to that found in recent issues of the Bulletin. First drafts of manuscripts should be typewritten double spaced and can be sent to any of the editors. After the manuscript has been reviewed and accepted, the author will be provided with a page format with which to prepare a single-spaced camera-ready copy of the manuscript. COORDINATING EDITOR william T. Boykins, Jr. Kay Clark-Bourne Kasandra D. Brockington Ian G. Macintyre ASSISTANTS National Museum of Natural History MRC-121 Smithsonian Institution PO Box 37012 Washington, DC 20013-7012 E-mail: macintyre.ian@nmnh.si.edu EDITORIAL BOARD Stephen D. Cairns Brian F. Kensley Mark M. Littler Wayne N. Mathis Jeffrey T. Williams Joshua I. Tracey, Jr. Warren L. Wagner Roger B. Clapp (MRC-163) National Museum of Natural History (MRC-163) (Insert appropriate MRC code) (MRC-166) Smithsonian Institution (MRC-169) Washington, D.C. 20560 (MRC-159) (MRC-137) (MRC-166) National Museum of Natural History National Biological Survey, MRC-111 Smithsonian Institution Washington, D.C. 20560 David R. Stoddart Department of Geography 501 Earth Sciences Building University of California Berkeley, CA 94720 Bernard M. Salvat Ecole Pratique des Hautes Etudes Labo . Biologie Marine et Malacologie Universite de Perpignan 66025 Perpignan Cedex, France PUBLICATIONS MANAGER A. Alan Burchell Smithsonian Institution Press STATUS OF CORAL REEFS IN THE WESTERN ATLANTIC: RESULTS OF INITIAL SURVEYS, ATLANTIC AND GULF RAPID REEF ASSESSMENT (AGRRA) PROGRAM 111 NO. 496 PAGES FOREWORD Robert N. Ginsburg and Judith C. Lang vii CAVEATS FOR THE AGRRA “INITIAL RESULTS” VOLUME Judith C. Lang xv SYNTHESIS OF CORAL REEF HEALTH INDICATORS FOR THE WESTERN ATLANTIC: RESULTS OF THE AGRRA PROGRAM (1997-2000) Philip A. Kramer (On behalf of the AGRRA contributors to this volume) 1 SURVEYS Bahamas A rapid assessment of coral reefs near Hopetown, Abaco Islands, Bahamas (stony corals and algae) Joshua S. Feingold, Susan L. Thornton, Kenneth W. Banks, Nancy J. Gasman, David Gilliam, Pamela Fletcher, and Christian Avila 58 Assessment of the Andros Island Reef System, Bahamas (Part 1 ; stony corals and algae) Philip A. Kramer, Patricia Richards Kramer, and Robert N. Ginsburg 76 Assessment of the Andros Island Reef System, Bahamas (Part 2: fishes) Philip A. Kramer,. Kenneth W. Marks, and Timothy L. Turnbull 100 Assessment of coral reefs off San Salvador Island, Bahamas (stony corals, algae and fish populations) Paulette M. Peckol, H. Allen Curran, Benjamin J. Greenstein, Emily Y. Floyd, and Martha L. Robbart 124 Belize Assessment of selected reef sites in northern and south-central Belize, including recovery from bleaching and hurricane disturbances (stony corals, algae and fish) Paulette M. Peckol, H. Allen Curran, Emily Y. Floyd, Martha L. Robbart, Benjamin J. Greenstein, and Kate L. Buckman 146 Brazil Rapid assessment of the Abrolhos Reefs, eastern Brazil (Part 1 : stony corals and algae) Ruy K.P. Kikuchi, Zelinda M.A.N. Ledo, Viviane Testa, Leo X. C. Dutra, and Saulo Spano 172 Rapid assessment of the Abrolhos reefs, eastern Brazil (Part 2: fish communities) Ruy K.P. Kikuchi, Zelinda M.A.N. Leao, Claudio L.S. Sampaio, and Marcelo D. Telles 188 iv Cayman Islands Status of coral reefs of Little Cayman, Grand Cayman and Cayman Brae, British West Indies, in 1999 and 2000 (Part 1; stony corals and algae) Carrie Manfrino, Bernhard Riegl, Jerome L. Hall, and Robert. Graifman 204 Status of coral reefs of Little Cayman and Grand Cayman, British West Indies, in 1999 (Part 2: fishes) Christy V. Pattengill-Semmens and Brice X. Semmens 226 Costa Rica A rapid assessment at Cahuita National Park, Costa Rica, 1999 (Part 1 : stony corals and algae) Ana C. Fonseca E. 248 A rapid assessment at Cahuita National Park, Costa Rica, 1999 (Part 2: reef fishes) Ana C. Fonseca E. and Carlos Gamboa 258 Cuba Rapid assessment of coral communities of Maria La Gorda, southeast Ensenada de Corrientes, Cuba (Part 1 : stony corals and algae) Pedro M. Alcolado, Beatrix Martinez-Daranas, Grisel Menendez-Macia, Rosa del Valle, Miguel Hernandez, and Tamara Garcia 268 Rapid assessment of coral communities of Maria La Gorda, southeast Ensenada de Corrientes, Cuba (Part 2: reef fishes) Rodolfo Claro and Karel Cantelar Ramos 278 Mexico Rapid assessment of Mexico’s Yucatan Reef in 1997 and 1999: pre- and post-1998 mass bleaching and Hurricane Mitch (stony corals, algae and fishes) Robert S. Steneck and Judith C. Lang 294 Condition of coral reef ecosystems in central-southern Quintana Roo, Mexico (Part 1 : stony corals and algae) Miguel A. Ruiz-Zdrate, Roberto C. Herndndez-Landa, Carlos Gonzdlez-Salas, Enrique Nuhez-Lara, andJ. Ernesto Arias-Gonzdlez 318 Condition of coral reef ecosystems in central-southern Quintana Roo, Mexico (Part 2: reef fish communities) Enrique Nuhez-Lara, Carlos Gonzdlez-Salas, Miguel A. Ruiz-Zdrate, Roberto Herndndez-Landa, and J. Ernesto Arias-Gonzdlez 338 Condition of selected reef sites in the Veracruz Reef System (stony corals and algae) Guillermo Horta-Puga 360 Netherlands ANTILLES Condition of coral reefs off less developed coastlines of Cura9ao (Part 1: stony corals and algae) Andrew W. Bruckner and Robin J. Bruckner 370 Condition of coral reefs off less developed coastlines of Cura9ao (Part 2: reef fishes) Andrew W. Bruckner and Robin J. Bruckner 394 A post-hurricane, rapid assessment of reefs in the Windward Netherlands Antilles (stony corals, algae and fishes) Kristi D. Klomp and David J. Kooistra 404 V St. Vincent A rapid assessment of the Horseshoe Reef, Tobago Cays Marine Park, St. Vincent, West Indies (stony corals, algae and fishes) Alice Deschamps, Andre Desrochers, and Kristi D. Klomp 438 TURKS AND Caicos Islands Assessment of the coral reefs of the Turks and Caicos Islands (Part 1 : stony corals and algae) Bernhard Riegl, Carrie Manfrino, Casey Hermoyian, Marilyn Brandt, and Kaho Hoshino 460 Assessment of the coral reefs of the Turks and Caicos Islands (Part 2: fish communities) Kaho Hoshino, Marilyn Brandt, Carrie Manfrino, Bernhard Riegl, and Sasha C.C. Steiner 480 UNITED States A rapid assessment of the Flower Garden Banks National Marine Sanctuary (stony corals, algae and fishes) Christy V. Pattengill-Semmens and Stephen R. Gittings 500 VENEZUELA Rapid assessment of coral reefs in the Archipielago de Los Roques National Park, Venezuela (Part 1 : stony corals and algae) Estrella Villamizar, Juan M. Posada, and Santiago Gomez 512 Rapid assessment of coral reefs in the Archipielago de Los Roques National Park, Venezuela (Part 2: fishes) Juan M. Posada, Estrella Villamizar, and Daniela Alvarado 530 Virgin Islands A rapid assessment of coral reefs in the Virgin Islands (Part 1; stony corals and algae) Richard S. Nemeth, Adam Quandt, Laurie Requa, J. Paige Rothenberger, and Marcia Taylor 544 A rapid assessment of coral reefs in the Virgin Islands (Part 2; fishes) Richard S. Nemeth, Leslie D. Whaylen, and Christy V. Pattengill-Semmens 566 ADDITIONAL AGRRA DATA Assessment tables for Abaco, Bahamas (fishes). Lighthouse, Belize (corals, algae, fishes) and Bonaire, Netherlands Antilles (corals, algae, fishes) Philip A. Kramer and Baerbel G. Bischof 590 SUPPLEMENTAL INEORMATION Condition of coral reef ecosystems in central-southern Quintana Roo, Mexico (Part 3: juvenile reef fishes) Carlos Gonzdlez-Salas, Enrique Nuhez-Lara, Miguel A. Ruiz-Zdrate, Roberto C. Herndndez-Landa, andJ. Ernesto Arias-Gonzdlez 598 APPENDIX ONE The Atlantic and Gulf Rapid Reef Assessment Protocols: Former Version 2.2 Patricia Richards Kramer and Judith C. Lang 611 APPENDIX TWO Fish biomass conversion equations Kenneth W. Marks and Kristi D. Klomp 625 VI LIST OF PLATES Prepared by Patricia Richards Kramer, Robert N. Ginsburg, and Judith C. Lang Plate 1. Location of AGRRA sites assessed as of mid 2003. xiv Plate 2. Caribbean reef crests: then and now. 56 Plate 3. Caribbean coral reefs: past and present baselines. 57 Plate 4. Reef coral condition. 123 Plate 5. Stony corals: partial mortality. 145 Plate 6. Stony corals: recent mortality. 247 Plate 7. Stony corals: old mortality. 267 Plate 8. Stony corals: disease. 359 Plate 9. Stony corals: bleaching. 393 Plate 10. Stony corals: predation. 403 Plate 11. Reef algae. 459 Plate 12. Diadema antillarum. 511 Plate 13. Reef fishes. 629 Plate 14. AGRRA methods. 630 FOREWORD BY ROBERT N. GINSBURG^ and JUDITH C. LANG^ INTRODUCTION The global decline in coral reefs during the last decades has provoked the most serious concerns about these remarkable ecosystems. Were they owing to a single worldwide cause, like influenza, plague or HIV in humans, the focus of efforts to understand and remedy the situation would be clear. Instead, the causes of declines as well as the nature of reefs vary significantly from region to region and within regions. Thus the urgent need is to assess the condition of reefs regionally with directly comparable quantitative observations rather than anecdotal reports. This volume contains the initial reports and their synthesis of a new approach to assessing the regional condition of coral reefs in the Western Atlantic Ocean developed under the Atlantic and Gulf Rapid Reef Assessment (AGRRA) program. It features rapid, multiscale assessments by teams of five-six trained observers for reefs of the Greater Caribbean, Gulf of Mexico and South Atlantic with the same method. Thus it becomes possible to assess many reefs spread over the entire region. The AGRRA protocols are focused on three key functional and structural elements of reef ecosystems: stony corals, fish and algae. The long-term goal of this region-wide effort is to provide, for the first time, a database suitable for comparative evaluation of current reef condition. This approach is similar to that which public health officials would use to make a rapid health assessment of villagers in remote areas. Analysis of the assessments provides norms of condition for some 30 key parameters of reef condition, such as the species identity, sizes and partial mortality of reef-building corals and the biomass of ecologically and commercially significant reef fishes. As demonstrated in the Synthesis of this volume, these initial norms, which are like those of human health assessments (blood pressure, pulse, reflexes), can be used to compare individual reefs at different spatial scales; between reefs; groups of reefs; subregions; or for most of the 20 reef areas in this volume. Even these initial results offer valuable background information for selecting protected areas or perturbed reefs in need of monitoring. While this volume was in preparation, additional AGGRA surveys have been carried out in several other major reef areas: Jamaica’s northern and western coasts; Cuba’s southern coast; the Caribbean reefs of Panama and the Florida Keys. The ’Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: rginsburg@rsmas.miami.edu ^ P.O. Box 539, Ophelia, VA, 22530. Email; JandL@rivnet.net Vlll completed regional assessment will expand and refine the initial regional interpretations presented in this volume, and produce an essential baseline with which to gauge future major changes disclosed by revisits. SPECIAL VALUE OF THESE INITIAL RESULTS Each of the assessments in this volume is based on data from multiple sites. The areas assessed extend from the northern Bahamas to the southernmost reefs of the Caribbean and from the windward Netherlands Antilles to the Gulf of Mexico. Some areas are of special interest because of significant anthropogenic impacts (e.g., Veracruz, Mexico, Costa Rica and some Virgin Islands sites). Others are notable for the relative scarcity (apart from fishing) of direct human impacts (e.g., Andros in the Bahamas, Flower Garden Banks in the NW Gulf of Mexico and Los Roques, Venezuela); all are affected by adverse regional- and global-scale changes. The inclusion of the southernmost reefs in the Atlantic off Brazil extends the range of areas examined and adds a coral community and reef structure quite different from that of the Greater Caribbean. These assessments were accomplished owing to the major efforts of about a hundred reef observers and team leaders from eight different countries who volunteered to spend long days of arduous diving to collect the basic data. The results of these first 20 AGRRA surveys demonstrate the promise of this approach for characterizing and comparing reef condition, for distinguishing between regional and local impacts, for recognizing the differences between acute and chronic stressor effects and for identifying candidate reef areas for protection or remediation. Some unanticipated but valuable dividends are the ancillary observations on the distribution of reefs, their general community structure, their geomorphology, and identification of localized threats to reefs. For example, the reef off Andros Island, some 2 1 7 km long, was so little known that our general characterization provides the most comprehensive information available on this second longest reef complex in the Western Atlantic. Kramer (this volume) has contributed an extensive Synthesis of the separate reports. The regional means cited below are based on this standardized analysis of the initial AGRRA dataset. Corals Our results provide quantitative verification of the decline of elkhorn coral (Acropora palmata) on reefs crests in Los Roques and the Tobago Cays (St. Vincent). Furthermore, what had been dense arrays of A. palmata on reef crests off Providencia in the Turks and Caicos Islands and Grand Cayman (Ginsburg, personal observations) and off Abaco (Bahamas) are now largely standing dead skeletons. Living staghorn corals (A. cervicornis) were rare in all areas, including those in which they are known to have occurred historically. That some of the die-offs of sensitive acroporid corals occurred in relatively remote areas reinforces earlier demonstrations that the declines are regional not local (e.g., Aronson and Precht, 2001). Surprisingly, healthy thickets of living elkhorn predominated on the shallow (1-3 m) crest along the Andros (Bahamas) reef tract. IX however, and signs of recovery and recruitment in localized areas elsewhere in the region are encouraging. The averages for living stony coral cover of 1 8 ± 1 0% in shallow reefs (<5 m) and 26 ± 13% in deeper reefs (10 ± 3m), which are based on 17 of the assessments, cannot be used as representative of the whole region for at least three reasons: a wide range of percent cover was found within and among reef areas; the means are not adjusted for the relative sizes of the different reef areas; and shallow reefs are under-represented in most areas. Partial mortality in stony corals represents the cumulative effects of diseases, overgrowth by algae and other epibionts, predation, bleaching, physical abrasion, etc. It can be subdivided by assessors on the basis of skeletal appearance into “recenf ’ (<~1 year) and “old” (>~1 year) mortality. As with coral cover, the regional mean of recent mortality of 4% of colony surfaces for both shallow and deep sites masks significant subregional variations. Recent mortality was well above the regional mean off Andros and in the western Caribbean (Belize, Yucatan) as a result of the cumulative effects of bleaching and diseases associated with the 1998 ENSO-related warming. The most severely affected taxa were Agaricia tenuifolia and species of the Montastraea annularis complex. Outbreaks of diseases in M. annularis and M faveolata were additionally noted off Cura9ao, the Cayman Islands, Costa Rica and some of the Virgin Islands. Of special concern is the prevalence of diseases and bleaching-related mortality in the M annularis complex, a major contributor to reef framework throughout the Greater Caribbean. Algae and Fishes Although much attention has been given to the current prominence of macroalgae on many wider Caribbean reefs, in terms of relative abundance, turf algae generally predominated in the AGRRA assessments. In deeper sites, however, elevated values for both macroalgal relative abundance and macroalgal index (a proxy for its biomass) were found throughout the Bahamas (Abaco, Andros and San Salvador) and in Maria la Gorda (Cuba). Herbivores affect the types, abundance and biomass of algae found on reefs. Herbivorous fishes [surgeonfishes (acanthurids), parrotfishes (scarids) >5 cm, and the damselfish Microspathodon chrysurus] averaged -30/100 m^ overall on deeper sites in the 1 7 areas having comparable data. Their density was not related to the relative abundance and index values of macroalgae for the region as a whole. Moreover, the long- spined urchin (Diadema antillarum), formerly a key herbivore, was too scarce (regionally <3/100 m ) to have any significant effect on algal distribution patterns. The total density of reef-associated AGRRA fishes (primarily the “AGRRA herbivores” plus commercially important predators) was nearly twice as high in shallow (85/100 m ) as in deeper (49/100 m ) sites for the 17 assessments mentioned above. Large-sized parrotfishes, seen mostly in the southern Caribbean, were rare. The overall mean density of large-sized groupers (serranids) and snappers (lutjanids) averaged <1/1 OOm^. The scarcity of large fishes is an indication that, regardless of location, legal designation, or local fishing regulations, the entire region has been overharvested at least for these species. X Synopsis Quantitative historical data with which to compare the present results are lacking for reefs in most of the assessed areas. Where prior information of some form exists, it is clear that only the relatively remote Flower Garden Banks have remained essentially unchanged in recent decades. Everywhere else (San Salvador in the Bahamas, Cayman Islands, Costa Rica, Los Roques, U.S. Virgin Islands, Yucatan) their condition has deteriorated. Kramer (this volume) shows how 13 of the 30 individual norms of condition can be used to establish a preliminary biotic health index for the 17 deeper assessments having comparable data. The four considered “better” (above average) are two in offshore locations (Flower Garden Banks and Los Roques), and two that are adjacent to small human populations (Bonaire and the windward Netherlands Antilles). In contrast, the six that were grouped in the “worse” (below average) category include sites in two sparsely populated biosphere reserves (Sian Ka’an, Mexico, and Guanahacabibes, Cuba), two of the Bahamian islands (Andros, Abaco), unprotected areas of the Yucatan, Mexico, and Costa Rica’s Cahuita National Park. Given these spatial patterns, no single type of threat, at any scale from localized anthropogenic inputs through regional overfishing and diseases to ENSO events and climate change, seems sufficient to explain the details of presumed or documented declines on reefs that have been assessed to date by the AGRRA protocols. DEVELOPMENT OF THE AGRRA PROGRAM The AGRRA program developed from insight afforded during the 1993 Colloquium and Forum on Global Aspects of Coral Reefs. This meeting, attended by some 1 20 scientists from 20 different countries, was an early attempt to consider the condition of reefs on a global scale. A major conclusion of the meeting highlighted the insufficiency of available information: “The database for evaluating the condition of the world’s reefs is quite inadequate on all counts,” (Ginsburg and Glynn, 1994). This conclusion was especially relevant to reefs of the Western Atlantic and Gulf of Mexico. Despite the extensive research on this region’s reefs beginning in the early 1900s, large areas had received little or no attention from reef scientists. [An exception is the ongoing international Caribbean Coastal Marine Productivity (CARICOMP) program, a pioneering effort to monitor reefs, sea grass communities and mangroves at a series of fixed localities around the region (KjerWe, 1998).] However, it had already been clear for about a decade that reef-building corals, most notably the branching elkhom and staghorn acroporid corals, were in serious decline at numerous sites in the Greater Caribbean. What was not clear in many areas was which other stony corals were affected. An initial effort to develop a standard method of assessing reef condition was focused solely on the reef-building coral community and followed the approach of Juan Manuel Diaz and his colleagues (Diaz et al., 1995) to record partial mortality of stony corals. The results demonstrated that a census of corals by species, size, and partial mortality could be done rapidly and provide useful comparisons of the condition of patch reef eorals in south Florida (Ginsburg et al., 1996; Ginsburg et al., 2001). XI The idea of a region-wide survey of Caribbean reefs was first discussed informally at the 1996 Reef Symposium in Panama. A major refocusing came later in 1996-1997 when Robert Steneck and Judy Lang proposed adding observations on algal functional groups, fish densities, herbivory, recruitment and the distinction between recent and old partial mortality, and then field-tested the protocols in three different geographic locations. Philip Kramer and Patricia Richards Kramer organized the first extensive field test of this expanded method in the Bahamas along the Andros reef tract in August 1997. It included quantitative fish assessments, following suggestions from Peter Sale, and roving diver surveys that were both conducted by Ken Marks. The positive results of all these trial applications encouraged us to post the protocols on the Internet and organize an international workshop that was held in Miami in June 1998. Eighty-one scientists from 19 different countries of the Greater Caribbean, Brazil, Canada, the United Kingdom, Austria and the Philippines participated in the five- day session to review and refine the prototype AGRRA protocols and plan their region- wide application. All the participants contributed to the development of the final product through lengthy discussions of the methods and a field trip to test the proposed revisions. The product of this workshop was Version 2 of the AGRRA protocols (see Appendix One, this volume) and most of the reports in this volume were based on that version. TRAINING AND APPLICATION OF THE PROTOCOL It was evident from the field trials before and during the 1 998 workshop that training was necessary to ensure the consistent application of the AGRRA protocols. Accordingly, Philip Kramer and Patricia Richards Kramer with Christy Pattengill- Semmens and Andrew Bruckner organized the first training workshop in Bonaire. Held in February 1999, for 1 1 participants, its success encouraged us to conduct a second workshop for reef scientists from Central America in Akumal, Quintana Roo in May, 1999. Philip Kramer, Patricia Richards Kramer, Andrew Bruckner and Elizabeth Fisher helped conduct this five-day bilingual session. The 25 participants came from Mexico (13), Belize (5), Honduras (3), Cuba (2), Costa Rica (1) and Colombia (1). ACKNOWLEDGMENTS Financial and In-kind Support The assessments reported in this volume were made possible with financial and in-kind support from several organizations and agencies. Major contributors of financial support to the project are The Bacardi Family Foundation, National Geographic Society’s Committee for Research and Exploration, The Ocean Research and Education Foundation, and World Bank/Netherlands Environmental Partnership Fund. Other donors include the Coral Reef Fund, Hachette-Filipacchi Foundation, Regional Co-ordinating Unit of the Caribbean Environment Programme of the United Nations Environment Programme, United States Department of the Interior, and The Henry Foundation. The National Fish and Wildlife Foundation and the National Center for Caribbean Coral Reef xii Research helped support the compilation of this volume. American Airlines provided in- kind support for travel. The Bonaire Marine Park and the Centro Ecologico Akumal, Mexico made facilities available for the training workshops. The authors in this volume acknowledge additional support for their individual assessments. Contributors Most the individual assessments reported in this volume required many long days of scuba diving and/or snorkeling, usually from small boats operating from shoreside bases, by participants who were all volunteers. Each assessment team consisted normally of six divers, four of whom made transects to evaluate corals and algae, and two who collected information on fish through belt transects and a roving diver census (for a description, see Appendix One, this volume). At the end of each day, team members then had to spend further hours transferring their results to a standard spreadsheet. Their commitment and hard work provided the basic data on which each report is based and for the Synthesis chapter. The leaders of these assessments naturally became the authors and coauthors of these reports. Their patience with, and responses to, the prolonged editorial and database construction process is much appreciated. The Synthesis chapter was made possible only through the use of an Access database. A major effort was required to ensure the consistency of the data entered in the database. Philip Kramer spent days correcting the initial entries through correspondence with observers. We are fortunate indeed that Kenneth Marks set up and maintained the database in such a way that queries could be answered in short order. Assembling this regional database was only possible owing to the generous sharing of basic data by the team leaders. We are grateful to all those who helped with the formatting and copy editing: Steven Eutz, Eric Eevorsen, Kimberly Rosen, Sara Fain, Barbie Bischof, Amit Hazra, and Karlissa Callwood at the Rosenstiel School of Marine and Atmospheric Science (RSMAS); Kay Clark-Boume and Ian MacIntyre at the Smithsonian Institution. We particularly wish to thank Patricia Richards Kramer for her constructive reviews and tireless assistance during the final preparation of the manuscripts and Eduardo Martinez for his many last-minute editorial corrections. Advisors From its beginnings at the 1996 Coral Reef Symposium, AGRRA has had the benefit of encouragement and constructive advice from many colleagues and outsiders in addition to those mentioned above. Among them are Pedro Alcolado (Instituto de Oceanologia, La Habana), Rolf Bak (Netherlands Institute for Sea Research), Barbara Best (United States Agency for International Development), Angel Braestrup (Flenry Foundation), Terry Done (Australian Institute of Marine Science), John Francis (National Geographic Society), James R. Gilbert and Lester Abberly (Showboats International), Peter Glynn, (University of Miami, RSMAS), Maria Hatziolos (World Bank), Thomas Hourigan (National Oceanic and Atmospheric Administration), William Kiene (Reef Check), Vladimir Kosmynin (Florida Department of Environmental Protection), Ken Lindeman (Environmental Defense), Ian MacIntyre (National Museum of Natural Xlll History), Ken Marks (Consultant), Kalli de Meyer (Coral), Steven Miller (National Undersea Research Center), Robert O’Brien (Bacardi Family Foundation), Kate Robinhawk (Centro Ecologico Akumal), Alessandra Vanzella-Khouri (Caribbean Environment Programme, United Nations Environment Programme) and Richie Williams (United States Geological Society). Regional Advisors We thank Pedro Alcolado (Cuba), Ernesto Arias (Mexico), Jorge Cortes (Costa Rica), Janet Gibson (Belize), and Zelinda Leao (Brazil) for their assistance with some of the field assessments. REFERENCES Aronson, R.B. and W.F. Precht 2001 . White-band disease and the changing face of Caribbean coral reefs. Hydrobiologia 460:25-38. Diaz, J.M., J. Garzon-Ferreira, and S. Zea 1 995. Los Arrecifes Coralinos de la Isla de San Andres, Colombia: Estado Actual y Perspectivas para su Conservacion. Academia Colombiana de Ciencias Exactas, Fisicas y Naturales. Editora Guadalupe Ltda, Colombia. 150 pp. Ginsburg, R.N., R.P.M. Bak, W.E. Kiene, E. Gischler, and V. Kosmynin 1996. Rapid assessment of reef condition using coral vitality. Reef Encounter 19:12-14. Ginsburg, R.N., E. Gischler, and W.E. Kiene 2001 . Partial mortality of massive reef-building corals: an index of patch reef condition, Florida Reef Tract. Bulletin of Marine Science 69:1 149-1 173. Ginsburg, R.N., and P.D. Glynn 1994. Summary of the colloquium and forum on global aspects of coral reefs: health, hazards and history. Pp. i-vii. In: R.N. Ginsburg (compiler). Global Aspects of Coral Reefs -Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Kjerfve, B. (ed.) 1998. CARICOMP — Caribbean Coral Reef, Seagrass and Mangrove Sites. Coastal region and small island papers 3. UNSECO, Paris, xiv + 347 pp. 90‘W 80"«/V 70W XIV 90W SOW 70W Plate 1. Location of all AGRRA sites assessed as of mid 2003. XV CAVEATS FOR THE AGRRA “INITIAL RESULTS” VOLUME BY JUDITH C. LANG' INTRODUCTION The Atlantic and Gulf Rapid Reef Assessment (AGRRA) collaboration is designed for small teams of trained observers to quickly collect relatively simple quantitative indicators of the condition and/or abundance of stony corals, benthic algal groups, and reef-associated fishes at specific depth intervals in certain zones of maximum reef development. Results of the early (August 1 997 to mid-2000) AGRRA assessments provide the focus of this volume. Coral reef ecosystems are so diverse, and their inhabitants engage in such intricate ecological relationships, that no rapid visual assessment technique can possibly provide in an unbiased manner all the information desired by scientists and resource managers for any given location. Comparisons among reefs are inherently constrained by numerous differences in physical environment, geomorphology, species composition, and proximity to direct human influences. Nevertheless, standardized application of the AGRRA methodology is facilitating multiscale spatial and temporal comparisons of key species, functional groups or guilds in the wider Caribbean (e.g., Ginsburg et al., 2000; Kramer, this volume). The purpose of this section is to alert readers to some of the special attributes of the AGRRA approach and some limitations in its initial application. QUALIFICATIONS General Considerations Versions. The AGRRA protocols have undergone several changes since their original posting in 1997 (see http://coral.aoml.noaa.gov/agra/method/methodhome.htm for the current version). Version 2, which is the basis for most of the research reported herein, is summarized in Appendix One (this volume). Given in the Methods section of each assessment paper are the particular version of the protocol that was used and any changes made in response to field conditions (or for any other reason). Sites. Site selection criteria, and the rationale employed when any sites were chosen for “strategic” purposes, are specified in the Methods. Nomenclature. The generic and specific names in the Methodology that were posted on the Internet and found in most of the papers in this volume are based on Foster (1987) for Stephanocoenia, Weil and Knowlton (1994) for the Montastraea annularis 'P.O. Box 539, Ophelia, VA, 22530. Email: JandL(@rivnet.net XVI species complex, and two publications of the American Fisheries Society (AFS): Cairns et al. (1991) for the remaining stony corals and Robbins et al. (1991) for fishes. The nomenclature and/or spelling of certain species differ from those given in the several editions of Paul Humann’s exemplary field guides that are widely used in the field by the AGRRA observers and/or from Eschmeyer’s (1998) revised Catalog of Fishes. Consistency. A degree of subjectivity is inherent in many of the decisions made when executing these protocols and subtle distinctions will occur as a function of the observer’s knowledge and level of experience. Consistency training to standardize the visual assessments followed by periodic reviews are two important components of the AGRRA methodology. Specific efforts to reduce observer bias among team members are described in the individual assessment papers. It must be admitted, however, that when divers and/or time become limited, in-water reviews are likely to be sacrificed. Inter- observer variability undoubtedly contributes to some of the larger variance values, especially in the means for individual assessment sites. Spatial coverage. During the time interval covered by the papers in this volume, several of the AGRRA geographic subregions were either poorly represented (e.g., Cuba, Lesser Antilles) or missing (e.g., Panama, Hispaniola) as was the entire Florida region. Due to various circumstances (e.g., funding, geography), some papers are limited in spatial coverage and/or in the number of assessed sites. Furthermore, the two high-relief habitats of particular interest (Acropora palmata in ~l-5 m, and fore reef or equivalent in ~8-15 m) are not present in all coral reef areas or, if present, sometimes could not be assessed for other reasons such as remoteness or weather. Temporal Coverage. Some of the individual assessments predate the 1998 ENSO while others either overlap with or postdate this extreme event, the effects of which were not experienced uniformly across the western Atlantic. Hurricanes (Georges, Mitch, Lenny) and outbreaks of disease also had nonrandom spatial and temporal distribution patterns profoundly affecting some of the assessed reef areas without influencing others. Particularly when recent partial mortality estimates of stony corals from different areas are compared, it is important to note the dates of the various assessments. Synthesis. In order to include all the data in the initial Synthesis, Kramer (this volume) has provisionally classified each site as either shallow (<5 m) or deep (>5 m) on the basis of its mean depth under the benthic transect lines with some resulting mixing of habitats and reef types. Each assessment has been treated as a separate unit and given equal weight in his analysis (Kramer, this volume). Hence its contribution is independent of the areal extent of the local reef system and of the numbers of assessed sites (or habitat types). In other words, the small (Costa Rica, Flower Garden Banks) and large (Cuba) areas, each with few assessed sites as of mid 2000, have been treated the same as the small (windward Netherlands Antilles) and large (Andros) areas for which a larger number of sites had been assessed (The corresponding numbers of habitats are three, one, two, numerous, and two, respectively.) XVI 1 As explained by Kramer (this volume), the specific datasets and methodology used for calculating site means in the Synthesis chapter differ from those employed in most of the individual assessment papers. Stony Corals Condition. The prevalence of bleaching, disease, predation, overgrowth, etc. are each expressed as a percentage of the surveyed population, i.e., all colonies of >10 cm (or >25 cm) maximum diameter that underlie the haphazardly placed 10-m transect lines. Observers vary in how much information they record as a function of time available and/or by their familiarity with these disturbances. Photographs and descriptions of the perturbations that commonly affect the wider Caribbean’s stony corals are now available at several web sites, in sets of laminated field cards, and in Bruckner (2002), yet there is no substitute for good, in-situ training. Given the variability of signs displayed by disturbed corals, however, even the most experienced observers are presently unable to reliably distinguish between the effects of certain diseases and certain predators, particularly during rapid “snapshots” like the AGRRA assessments. Therefore “absence of evidence” in some locations cannot be taken as necessarily indicating “evidence of absence.” For example, the AGRRA geologists have a tendency to report a higher proportion of stony corals that are “standing dead” (= completely dead with the colony still in growth position and recognizable at least to genus) than have the AGRRA biologists (P. Kramer, personal communication). Mortality. “Recent” and “old” mortality of stony corals is estimated as the percentages of their outward-facing surfaces that are dead when seen from above the colonies. Hence, average “partial-colony mortality” (or partial mortality) refers to the mean percent of tissue loss/colony and not to the percent of colonies with any (necessarily unspecified amount of) tissue loss. “Recent partial mortality” (after Diaz et al., 1995) in the AGRRA benthos protocol encompasses that percentage of the colony surface in which the skeleton is white and covered by a (necessarily thin) layer of algae or fine mud. “Old partial mortality” is used to describe the corresponding percentage in which the skeletal structures are no longer white and have either been lost or are covered by epibenthic organisms that are not easily removed. Standing dead corals are included in the calculations of mean values for old and total (= recent + old) partial mortality in all but one of the individual assessment papers; in the Synthesis they are excluded from the mean values of old partial mortality (Kramer, this volume). Attempts to “flesh out” these definitions and, in the absence of published data, to add putative temporal ranges to the definition of recent mortality, are given in the Methodology section of the AGRRA web site and by Bruckner and Bruckner (this volume), Kramer (this volume), and Steneck and Lang (this volume). Recently occurring mortality will be underestimated when: (a) exposed skeletal surfaces are quickly covered with sediment and/or algae (Fonseca, this volume); (b) turf algal-sediment mats expand at the expense of stony corals without creating any noticeable “recently dead” areas at their interfaces (Roy, personal communication); and (c) in the presence of superior spatial competitors (Deschamp et al., this volume) like the rapidly growing Trididemmm solidum (Bak et al., 1981) since the skeleton that is being overgrown is never exposed to view. XVlll Algae Functional algal groups are characterized by their abundance in 25 cm x 25 cm (= 0.0625 m ) quadrats with at least 80% coverage by any kind of benthic algae. The location of the quadrats is spatially limited to a 1-m radius of the 2-m marks on the transect lines. Although not a measure of algal cover, the abundance of each group on exposed substrata that are available to herbivores is provided by these data. As the identity of the functional groups that are assessed was changed between Versions 2 and 3 of the protocol, the usage of “relative abundance” has been restricted in this volume to the groups that were estimated in Versions 1 and 2 (i.e., macroalgae, turf algae and crustose coralline algae). Fishes The AGRRA benthos protocol is a novel collaborative creation (see the Forward and Appendix One, this volume) that is still being fine-tuned as we gain experience with its application in diverse geographical areas. In contrast, only minor adjustments have been implemented thus far with the fish belt transects (here restricted to ecologically important herbivores and commercially significant carnivores) and Roving Diver Technique (Schmitt and Sullivan, 1996). Both had been thoroughly tested for some years prior to their adoption for the AGRRA fish protocol. Hence their relative strengths and limitations are better understood (e.g.. Brock, 1954; Sale, 1980; Thresher and Gunn, 1986; Fowler, 1987; Schmitt et ah, 2002). For example, serranids (e.g., Pattengill- Semmens and Semmens, this volume) are generally underreported in the belt transects, especially on reefs with high structural complexity (Kramer, Marks and Turnbull, this volume). Also underestimated in belt transects are roving schools of scarids or acanthurids (Nemeth et ah, this volume). While the Roving Diver Technique provides a relatively rapid quantification of reef fish assemblage, longer search times or a larger number of searches than are appropriate for rapid assessments would be needed to fully estimate species richness (Nemeth et ah, this volume; Marks, personal communication; Semmens, personal communication). AFFIRMATION Caveats notwithstanding, our understanding of reef condition in the western Atlantic is enhanced as a result of the early (August 1997 to mid 2000) AGRRA efforts reported in this volume. Some important geographical gaps have been filled during subsequent assessments: northern and western Jamaica; southwestern and south-central Cuba; Bocas del Toro and western Kuna Yala, Panama; and Upper and Lower Keys, Florida. The data added from these (and remaining as-yet unvisited) areas is certain to modify some of the initial conclusions presented in this volume. Pending the outcome of their analysis we anticipate being able to provide a more complete accounting of the overall status of the coral reefs in the Intra-Americas Seas and Brazil. XIX ACKNOWLEDGMENTS Robert Ginsburg is thanked for the probing discussions that precipitated this cautionary preamble. Comprehensive reviews by Robert Ginsburg and Patricia Richards Kramer have clarified its execution. The specific comments of Ana Fonseca Escalante, Kristi D. Klomp, Kenneth Marks, Richard Nemeth, Paulette Peckol, Juan Posado, Roshan Roy, Miguel Ruiz, Brice Semmens, and Robert Steneck are also greatly appreciated. REFERENCES Bak, R.P.M., J. Sybesma, and F.C. van Duyl 1981. The ecology of the tropical compound ascidian Trididemnum solidum.ll. Abundance, growth and survival. Marine Ecology Progress Series 6:43-52. Brock, V.E. 1 954. A preliminary report on a method of estimating reef fish populations. Journal of Wildlife Management 18:297-308. Bruckner, A. 2002. Appendix II. Coral health and mortality. Recognizing the signs of coral diseases and predators. Pp. 240-271. In: P. Humann and N. DeLoach (eds.). Reef Coral Identification. Florida, Caribbean, Bahamas. New World Publication, Jacksonville, FI. Cairns, S.D., D.R. Calder, A. Brinkmann-Voss, C.B. Castro, P.R. Pugh, C.E. Cutress, W.C. Jaap, D.G. Fautin, R.J. Larson, G.R. Harbison, M.N. Arai, and D.M. Opresko 1991. Common and Scientific Names of Aquatic Invertebrates from the United States and Canada. Cnidaria and Ctenophora. American Fisheries Society Special Publication 22. Bethesda, Maryland. 75 pp. Diaz, J.M., J. Garzon-Ferreira, and S. Zea 1 995. Los Arrecifes Coralinos de la Isla de San Andres, Colombia: Estado Actual y Perspectivas para su Conservacion. Academia Colombiana de Ciencias Exactas, Fisicas y Naturales. Editora Guadalupe Ltda, Colombia. 150 pp. Eschmeyer,W.N. (editor) 1998. Catalog of Fishes. Special Publication No. 1 of the Center for Biodiversity Research and Information, California Academy of Science, San Francisco, CA. Volumes 1-3, 2905 pp. Foster, A.B. 1987. Neogene paleontology in the northern Dominican Republic. 4. The Genus Stephanocoenia (Anthozoa: Scleractinia: Astroeoeniidae). Bulletins of American Paleontology 93:5-22. Fowler, A.J. 1987. The development of sampling strategies for population studies of coral reef fishes. A case study. Coral Reefs 6:49-58. XX Ginsburg, R., P. Alcolado, E. Arias, A. Bruckner, R. Claro, A. Curran, A. Deschamps, Fonseca, J. Feingold, C. Garcia-Saez, D. Gilliam, S.D. Gittings, A. Glasspool, G. Horta- Puga, K. Klomp, P.A. Kramer, P.R. Kramer, Z. Leao, J. Lang,, C. Manfrino, R. Nemeth, C. Pattengill Semmens, P. Peckol, J. Posada, B. Riegl, J. Robinson, P. Sale, R. Steneck, J. Vargas, and E. Villamizar 2000. Status of Caribbean Reefs: Initial results from the Atlantic and Gulf Rapid Reef Assessment (AGRRA) Program. Proceedings of the Ninth International. Coral Reef Symposium Abstract, p. 21 1 . Robbins, C.R. and R.M. Bailey, C.E. Bond, J.R. Brooker, E.A. Lachner, R.N. Lea, and W.B. Scott 1991. Common and Scientific Names of Fishes from the United States and Canada. Fifth Edition. American Fisheries Society Special Publication 20. Bethesda, Maryland. 183 pp. Sale, P.F. 1980. The ecology of fishes on coral reefs. Oceanography and Marine Biology Annual Reviews 18:367-421. Schmitt, E.F., T.D. Sluka, and K.M. Sullivan-Sealy 2002. Evaluating the use of roving diver and transect surveys to assess the coral reef assemblages off southeastern Flispaniola. Coral Reefs 21:216-223. Schmitt, E.F., and K.M. Sullivan 1996. Analysis of a volunteer method for collecting fish presence and abundance data in the Florida Keys. Bulletin of Marine Science 59:404-416. Thresher, R.E., and J.S. Gunn 1986. Comparative analysis of visual census techniques for highly mobile, reef- associated piscivores (Centrachidae). Environmental Biology of Fishes 17:93-116. Weil, E., and N. Knowlton 1994. A multi-character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bulletin of Marine Science 55:151-175. Atlantic and Gulf of Mexico Region Figure 1 . Map of western Atlantic showing AGRRA regional division boundaries and site locations for surveys reported on in this volume. SYNTHESIS OF CORAL REEF HEALTH INDICATORS FOR THE WESTERN ATLANTIC: RESULTS OF THE AGRRA PROGRAM (1997-2000) BY PHILIP A. KRAMER* (On behalf of the AGRRA contributors to this volume) ABSTRACT The Atlantic and Gulf Rapid Reef Assessment (AGRRA) sampling strategy is designed to collect both descriptive and quantitative information for a large number of reef vitality indicators over large spatial scales. AGRRA assessments conducted between 1998 and 2000 across a spectrum of western Atlantic reefs with different histories of disturbance, environmental conditions, and fishing pressure were examined to reveal means and variances for 15 indicators. Twenty surveys were compiled into a database containing a total of 302 benthic sites (249 deep, 53 shallow), 2,337 benthic transects, 14,000 quadrats, 22,553 stony corals. Seventeen surveys contained comparable fish data for a total of 247 fish sites (206 deep, 41 shallow), 2,488 fish transects, and 71,102 fishes. Shallow (< 5 m) reefs were dominated by A. palmata, a good proportion of which was standing dead, while deep (>5m) reefs were nearly always dominated by the Montastraea annularis species complex. Fish communities were dominated by acanthurids and scarids with seranids making up less than 1% of the fish seen on shallow reefs and 4% on deep reefs. AGRRA benthic and fish indicators on deep reefs showed the highest variation at the smallest spatial scale (~<0.1 km), with recent mortality and macroalgal canopy height displaying the largest area and subregional scale (~1-100 km) variation. A mean live coral cover of 26% for the 20 survey areas was determined for the deep sites. Significant bleaching and disease-induced mortality of stony corals associated with the 1 998 (El Nino-Southern Oscillation) ENSO event were most apparent in the western Caribbean and Bahamas subregions and the Montastraea annularis complex was the most heavily impacted. The overall low ntimber of sightings for larger-bodied groupers and snappers (~< 1/100 m^) as a whole suggest that the entire region is overfished for many of these more heavily targeted species. More remote reefs showed as much evidence of reef degradation as reefs more proximal to human coastal development. Characterizing present-day reef condition across the region is a complex problem since there are likely multiple sources of stress operating over several spatial and temporal scales. Not withstanding the many ' Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: pkramer@rsmas.miami.edu 2 limitations of this analysis, the value of making multiple observations across multiple spatial scales that can approximate the “normal” state for the region today is still very high. INTRODUCTION Coral reefs in the western Atlantic have undergone massive changes over the past several decades from coral-dominated to algal-dominated states. Widespread impacts such as the 1983 die-off of Diadema antillarum and severe overfishing have disrupted herbivory processes. Coral diseases, like the white-band disease epizootic (Green and Bruckner, 2000) and bleaching (Wilkinson, 2000) have decimated previously healthy coral populations. Our understanding of these and other decadal-scale changes comes largely from a limited number of descriptive studies where long-term monitoring has been conducted. Some of the most frequently cited examples are those from Jamaica (Hughes, 1994), Florida Keys (Dustan and Halas, 1987; Porter and Meir, 1992), Cura9ao (Bak and Nieuwland, 1995); and Belize (Aronson and Precht, 1997). With the exception of the Caribbean Coastal Marine Productivity (CARICOMP) program (Kjerfve et al., 1998) and the fish surveys of the Reef Environmental and Education Foundation (REEF) (www.reeforg), few studies in the western Atlantic (Caribbean, Gulf of Mexico, Florida, Bahamas, and Brazil) have examined the condition of coral reefs over large spatial scales (100s- 1000s km). Moreover, most studies have focused on localized impacts and used diverse sampling methods making statistical comparisons on larger spatial scales impossible. Establishing regional patterns of reef condition is essential for characterizing the extent and severity of decline and developing hypotheses about the causes of decline (Ginsburg and Glynn, 1994). AGRRA was conceived to provide a “snapshof ’ characterization of a number of structurally or functionally important benthic and fish indicators on western Atlantic coral reefs. When applied synoptically to the entire region, results can be used to develop a biotic index of relative health or condition. The concept of evaluating “ecosystem health” is a rapidly emerging science for which a number of different definitions and approaches have been suggested (e.g.. Costanza, 1992; Rapport et al., 1998, 1999). Because health involves the response of structural and functional components of the ecosystem, most approaches have used more than a single indicator. Some of these include naturalness, normality, productivity, organization (species diversity and complexity of interactions), and resilience (Coates et al., 2002). Normality in itself has been shown to be particularly useful since it provides guidelines for the range of system states. The AGRRA approach relies on normality as its principal measure of condition and these norms are meant to represent a baseline for the region. The methodology and indices, which were developed with the advice of specialists and based on current scientific understanding of coral reef dynamics as well as standard monitoring methodologies, are summarized below. 3 Coral Condition The condition of the principal scleractinian and hydrozoan corals that contribute most to the construction and maintenance of the three-dimensional framework is critical for determining the long-term integrity of the reef ecosystems (Dustan and Halas, 1987; Done, 1997). Species composition, colony size, mortality, recruitment, disease, bleaching, predation, coral cover, etc., are examined over large spatial scales in the AGRRA assessment. Estimates of colony size provide information on rugosity, architectural complexity, and an approximation of colony age (Hughes and Jackson, 1980, 1985). Visual estimates of the partial mortality of colonies (hereafter partial mortality) are used to distinguish between the amount of moribund tissue considered recently dead (about 1 year) and long dead (more than 1 year). The amount of “recent mortality” indicates current impacts while “old mortality” is primarily an integration of mortality over longer time scales. The AGRRA data, within certain limits, can be used to calculate size frequency distributions (Bak and Meesters, 1998) as well as identify mortality patterns related to size and/or species identity. Given current knowledge it is hypothesized that large percentages of recent partial mortality are a signal of decline yet some level of old mortality is expected, at least in the larger corals (Hughes and Connell, 1999). By examining coral condition indices on many types of reefs throughout the region, patterns should emerge to establish and help formulate hypotheses on causes. Algae There has been a noticeable change on many reefs around the wider Caribbean from coral-dominated communities to those dominated by macroalgae (e.g.. Done, 1992; Hughes, 1994; Hallock et al., 1993; Dustan and Halas, 1987; Lewis, 1986; Steneck and Detheir, 1994; McClanahan and Muthiga, 1998). The causes for these shifts have been attributed, in part, to a loss of key herbivorous fishes and sea urchins, particularly Diadema antillarum. Diadema affects coral reef structure and composition, including algal composition and abundance, by competition with other grazers, particularly certain fishes, and by erosion of coral skeletons (e.g., Steneck and Dethier, 1994; Roberts, 1995; McClanahan and Muthiga, 1998). The dramatic decline of Diadema that occurred after its 1983 die-off (Lessios et al., 1984) led to the dominance of many reefs in Jamaica by fleshy and calcareous macro algae, increased coral mortality and decrease in coral recruitment (Hughes, 1994). Recent reports have suggested the abundance of Diadema has been increasing in localized areas but little regional information is available on its recovery and subsequent influence on the condition of reefs (but see Edmunds and Carpenter, 2001). The objective of assessing algae is to quantify the relative abundance of several key functional groups (crustose corallines, macroalgae and, initially, turf algae), and to relate these abundances with herbivorous fish biomass and coral condition. Given that reefs in decline often have high fleshy (noncalcified) macroalgal biomass, sometimes accompanied by a high biomass of Halimeda, it is expected that reefs with a low macroalgahcrustose coralline ratio, or a low macroalgal index (= macroalgal abundance x 4 macroalgal height) are more biologically intact than those with a high ratio (after Steneck and Dethier, 1 994). Fishes Reef fish, as predators or grazers, play important roles in the community dynamics of coral reefs through their interactions with corals, algae and other herbivores (Roberts, 1995). Fish communities respond to disturbance in various ways depending on the type and degree of perturbation. Various combinations of commercial, subsistence and recreational fishing, particularly of herbivores, constitute some of the most widespread and greatest anthropogenic impacts on coral reefs (Roberts, 1995). In particular, the loss of certain indicator species and guilds causes both direct and indirect shifts in fish community structure as well as in other components of reef ecosystems (Munro and Williams, 1985; McClanahan and Muthiga, 1988; Hughes, 1993, 1994). Disruption in the balance of reef fish assemblages can result in decreased coral cover and increased algal abundance (Roberts, 1995, 1997; McClanahan et al., 1996). Yet only a few studies have examined the response of fish communities to degradation or changes in habitat structure and composition (Jones and Syms, 1998). Whereas visual fish censuses have been conducted throughout the Caribbean and Gulf of Mexico, relatively few are comparable interregionally because of the various methodologies used (Sale, 1998). The AGRRA approach includes two distinct assessment methods that provide complementary“snapshots” of fishes at a given site although they do not fully account for the daily, tidal, and seasonal changes known to occur in reef habitats (Ault and Johnson, 1998; Bellwood, 1988; Willis, 2001). While subtle differences among fish assemblages can be difficult to detect with this level of sampling, robust patterns across large number of assessments can be revealed. It is expected that areas near human populations will have lower abundance of fishes, particularly commercially significant species. In this paper I have synthesized the AGRRA data on coral reef condition collected between 1 998 and 2000 by a network of scientists from 20 distinct locations in the western Atlantic extending over 1,000 km (Table 1). The areas presented in this synthesis include a wide spectrum of situations with respect to history of disturbance, environmental conditions, and fishing pressure from humans. The principal goals of this synthesis are to examine over various spatial scales (0.1-100 km): 1 ) variability of coral condition (mortality, recruitment, disease, damage), 2) the relative abundances of major algal functional groups and factors contributing to any differences in macroalgal index, 3) patterns of spatial variability of the abundance and size of targeted fish species or key guilds such as herbivores to evaluate the degree of overfishing, and 4) the integration of indices into a biotic “reef health index” that can be used in an exploratory way to examine patterns and form hypotheses for future experimental testing. This synthesis is unique in providing the first regional perspective on coral reef condition in the western Atlantic that is based on multiple indicators and in its 5 examination of spatial variation and trends of reef condition on multiple spatial scales. The results summarized here originate entirely from the data generously contributed by the authors of this volume to the AGRRA database and may differ slightly from that reported in their papers because of differences in the way the data have been analyzed. The reader is referred to individual papers in this volume for more specific information about each of the assessment areas. METHODS Sampling Design A map of the western Atlantic showing AGRRA regional division boundaries and geographic location (indicated by identification number) for the 20 assessments reported on in this volume is shown in Figure 1. All data submitted by individual AGRRA teams in the form of preformated spreadsheets were checked for errors and standardized to uniform species codes. The timing of the assessments spans a three-year period, many of which (~12) were conducted during the wet-season months (June-August) in much of the wider Caribbean. To synthesize results from the many assessments, an Access database was developed that allows users to manage large quantities of data, make statistical comparisons across various spatial scales, and provide this information to other AGRRA scientists, resource managers and interested researchers. A version of the database (Version 1.1) covering the period from 1997-2000 is currently available and newer versions will be released in the future. Fish data collected using the Roving Diver Technique are not discussed here as they are not part of the AGRRA database but instead are housed in the REEF database (www.reef org). A synopsis of the AGRRA methodology is provided in Appendix One (this volume). The majority of the AGRRA teams used Version 2 of the protocol. Conceptually, the AGRRA sampling design is based on principles of stratified two-stage sampling (Cochrane, 1977). A hierarchical multiscale sampling approach and associated spatial scales are shown in Figure 2. The sampling domain, defined as the western Atlantic region, is subdivided spatially into subregions, areas, reefs, and sites. Eight subregions are recognized in the western Atlantic: Gulf of Mexico, Bahamas, western Caribbean, central Caribbean, eastern Caribbean, southern Caribbean, Brazil, and Florida. AGRRA data have been collected from each of these subregions except Florida, for which an assessment is scheduled in summer 2003. Further stratification of reef types can be accomplished using a combination of depth, cross-shelf position, and geomorphology to delineate areas of similarity over any specified geographic area. One aim of AGRRA sampling is to select a series of “typical” sites that are representative of the geographic area. For the purpose of this synthesis, however, sites were only distinguished based on the mean benthic transect depth into “shallow” (<5 m, mostly reef crests and patch reefs) and “deep” (> 5m, mostly fore-reef slopes) categories. The appropriate sampling effort needed to characterize sites, areas, and sub- regions can vary from location to location depending on the spatial variation of coral. 6 AGRRA hierarchical sampling design Number Spatial unit name Approximate Spatial scale 1 Western Atlantic (including Gulf of Mexico) >1000 km 8 (see map) Regions 1000 km # varies by Region Sub regions 100 km # varies by Subregion Areas 10 km typically 2-5 within an Area Reef 1 km Site 0.1 km / \ Shallow (<5 m) Deep (> 5 m) reef crest back reef patch reef other Fore reef Bank reef other Figure 2. Hierarchical sampling scheme and associated spatial scales for AGRRA surveys. Sites are also classified hierarchically based on depth, geomorphology, relief, and additional modifiers. For the purpose of this synthesis, sites are only distinguished based on depth into shallow (<5 m) and deep (>5 m). 7 algal and fish indicators. Often AGRRA surveys are the first to be conducted in an area and/or little or no suitable baseline data exist to determine the appropriate level of sampling that would be needed to characterize the area fully. Furthermore, each indicator will have a separate distribution and desired degree of precision. A power analysis was conducted for several of the AGRRA indicators to examine the tradeoff between sampling effort and precision at both the site and subregional spatial scales. To assess how many transects were needed to adequately estimate site means for each indicator, within-site variance as a function of sampling effort was examined at one site in each of four areas. Data Analysis The database used for this synthesis contained a total of 302 benthic sites (249 deep, 53 shallow), 2,337 benthic transects, 14,000 quadrats, 22,553 corals, 247 (206 deep, 41 shallow) fish sites, 2,488 fish transects, and 71,102 fishes. For this initial analysis, 1 divided the sites into either shallow (<5 m) or deep (>5 m) depth categories based on the mean depth of the benthic transects, whereas sites were categorized by depth ranges or by mean depth of the habitat in some of the individual assessments in this volume. In several cases (e.g., Cura9ao and the Virgin Islands) the same geographic site was assessed at different times. 1 treated these sites as independent samples whereas in the individual assessment papers they are grouped together as a single site. Not all results described in the individual assessments were included in the synthesis either due to inconsistencies in the way the methodology was implemented or because of missing data for individual transects or entire sites. The cylindrical (volumetric) fish counts used in Belize and San Salvador, Bahamas (see both papers by Peckol et al., this volume) could not be directly compared to the belt transect data. The 1997 Andros, Bahamas survey followed an early version of the protocol and differed enough from later versions that only data collected during 1998 were used. To equalize sampling effort, results in the synthesis have only incorporated a maximum of the first 10 transects/site in calculated averages for either fish or benthic parameters, despite larger sample numbers for some sites. Statistical analysis was performed using STATISTICA software. Version 6.0 (StatSoft, Inc., 2002). The percent live coral cover, percent partial-colony mortality (recent, old, total), standing dead (i.e., completely dead and still in original growth position, and calculated as a percentage of the total), mean colony size, prevalence of disease and bleaching, and relative algal abundance were calculated and summarized. Only data from corals that were >25cm in maximum diameter was included. I chose to remove standing dead colonies from old partial mortality averages; thus, averages in the synthesis should be lower than those in the individual papers for sites with significant amounts of standing dead coral. The macroalgal index, a proxy for macroalgal biomass, was calculated as % relative macroalgal abundance (as approximated in Versions 1 and 2 of the AGRRA protocol) x macroalgal canopy height. Of the 75 species of fishes in the AGRRA belt transect list (see Appendix One, this volume), a total of 18 species of herbivores, both territorial and non- territorial, were analyzed (all acanthurids, scarids >5 cm, Microspathodon chrysurus (yellowtail damselfish) and black durgon {Melichthys niger). Free-swimming predators 8 and sedentary predators were classified as carnivores (all lutjanids, all serranids, and Sphyraena barracuda) for a total of 24 species. Fish densities (#/100 m^) were calculated for each area for all 75 species and included all sightings except for scarids and haemulids where only the >5 cm sightings were used. Size information was used to calculate biomass for each fish species using the standardized conversion equations shown in Appendix Two (this volume). Regional norms for different indicators were calculated by averaging reported survey values without weighting by the spatial extent of reefs or number of sites conducted within a given survey area. Several parameters were analyzed by students t-test and by 1-way and 2-way Analysis of Variance (ANOVA). The Plymouth Marine Laboratory’s PRIMER software, Version 5 (Clarke and Gorley, 2001), was used to examine similarity among different indicators with either Bray Curtis similarity or two- dimensional multidimensional scaling (MDS) ordination. To examine relationships between fish data and benthic habitat variables, regression analysis was used to analyze (1) mean herbivore density and mean macroalgal index for 1 7 assessments and (2) algal canopy height and density of the >25 cm corals. To examine the relationship of coral condition with respect to anthropogenic threats, survey areas were classified into one of three threat categories (high, medium, low) based on Bryant et al.’s (1998) global threat analysis (Table 1). The modeled threat layer of “overexploitation of marine resources,” which is based primarily on proximity of a reef to coastal settlements, was used to examine relationships of fish density and biomass to presumed fishing pressure. RESULTS Results in the Synthesis differ somewhat from those reported in the individual papers in this volume due to differences in how sites were defined or stratified, in the selection criteria for calculating site means and, for stony corals, in the definition of old partial mortality (see Methods). Sampling Effort The characteristics of the sampling effort for the 20 assessments in this volume are summarized in Table 1. Each assessment is identified by its unique identification number (ID#) in both figures and tables. At least some deep (>5 m) habitats were assessed by all teams but shallow (<5 m) reefs were examined extensively only in Andros (1D#2) and the Abrolhos, Brazil (ID#20). For this reason, most of the spatial comparisons in this synthesis are restricted to deep (>5 m) habitats. A detailed summary of the mean sampling effort for these deep sites in each assessment is given in Table 2. Sampling size (as number of transects) versus standard error for live coral cover and coral density from four different areas of the western Atlantic [Virgin Islands (ID# 13), Veracruz, Mexico (ID#6), Andros (ID#2), windward Netherlands Antilles (ID#14)] are shown in Figure 3. For both parameters, the error was reduced most substantially within the first six transects after which only small improvements were 9 Figure 3. Sampling size (number of transects) versus standard error for four sites in different areas of the western Atlantic for live coral cover and coral density. For all parameters, the error is reduced most significantly within the first six transects, after which only small improvements are observed with increasing number of transects. 10 obsei'ved with increasing sample size. Figure 4 provides a summary of sampling size for several AGRRA benthic indicators versus standard error. When percent partial-colony mortality (recent and old) and coral diameter (as number of corals) were compared for four sites [Cayman (ID# 12), Cura9ao (ID# 18), Belize (ID# 10), and Andros (ID#2)], error was reduced most dramatically within the first 50 corals sampled. Algal relative abundance (macroalgae, crustose corallines and turfs) and macroalgal canopy height data were compared (as number of quadrats) for Veracruz, Abrolhos (ID# 20), Cayman, and Andros, with the standard error most reduced within the first 40 quadrats sampled. In nearly all of the fish belt-transect surveys, an average of 600 m was assessed at each site with the exceptions being Akuma/Xcalak (ID#8) and Cura9ao (ID# 18), each with less than 200 m^. In both Maria la Gorda, Cuba (ID#1 1) and the Yucatan (ID#7), 50 X 2 m transects (a total of six per site) were used in accordance with an early version of the fish protocol, whereas 10 x 2 m transects (30/site) were employed in the small reefs of the Abrolhos (ID#20). The 14 remaining teams each performed 30 x 2 m transects (10/site). Stony Corals Composition and abundance. The 22,553 colonies sampled in the 20 surveys combined included 37 seleractinian and two hydrozoan species. The majority of these stony corals were assessed in deep (n=18,913 at >5 m) rather than shallow (n=3,640 at <5 m) sites. Of these, 1,122 colonies (5%) were surveyed in the Abrollos (ID#20), where the endemic Brazilian scleractinain, Mussismilia braziliensis, was the dominant coral. For the remaining corals in the wider Caribbean excepting Veraeruz (ID#6) where they were not assessed, standing dead colonies constituted 13.5% of the colonies at shallow sites and 2.5% at deep sites. The relative species composition, expressed as a percent of the total of corals (excluding all standing dead colonies and all the Brazilian corals) is shown in Figure 5A,B. Numerically the most abundant species in shallow reefs were Acropora palmata (27% of total), Montastraea annularis (19% of total), and Porites spp. (7% P. astreoides, 6% P. porites of total) (Fig. 5A). Live Acropora palmata was the dominant coral at 60% of all shallow reef-crest sites. Taxa that are usually dominant on shallow Caribbean reefs, such as Agaricia tenuifolia and Millepora spp. were less abundant in the AGRRA regional dataset. For shallow reefs, there was especially high between-survey species variability caused, in part, by differenees in the type of shallow reef assessed (reef crest, back reef, patch reef-see individual papers in this volume and Table 1). The high number of Acropora palmata reflects the relatively large number (Table 1) of reef crests sites assessed off Andros (ID#2) where this coral was common. Live Acropora palmata was the dominant coral (>l/3 of coral population) at 17 of 23 shallow reef-crest sites summarized in this volume. Deep reefs displayed less species variability among sites since most assessments were conducted in fore reefs [exceptions being bank reefs surveyed in the Flower Garden Banks, Gulf of Mexico (ID#5) and Mouchoir Bank in the windward Netherlands Antilles (ID# 14); see Table 1 and papers in this volume]. Montastraea spp. dominated most deep 11 cc ■D c 3 CO LU "E <5 *a c 3 CO 15 10 LU ■D <5 1 5 TO CO LU ■o TO T3 C iS CO Quacirats samplecJ (#) Figure 4. Sampling size (number of corals and quadrats) for several AGRRA benthic indicators versus standard error for four sites in different areas of the western Atlantic. For corals, error is reduced most dramatically within the first 50 corals sampled. For quadrats, the error is reduced most significantly within the first 40 quadrats sampled. 12 A Shallow (<5 m) stony corals A. cervicornis ■i'fo M. faveolata n = 2,712 B Deep (> 5 m) stony corals A. palmata 1 o/^ Figure 5. Composition of all stony corals (25 cm) from all assessments except Brazil (ID #20) combined for (A) shallow sites (5 m), and (B) deep sites (>5m). Includes only corals sampled within a maximum of 1 0 transects per site. 13 reef assemblages as follows: M. annularis (19% of total), M faveolata (13% of total), M. cavernosa (9% of total) and M franksi (8% of total). Other commonly observed taxa on fore reefs included Agaricia spp. (9% of total), Siderastrea siderea (7% of total), Diploria strigosa (6% of total), Porites spp. {P. astreoides 6%, P. porites 4% of total), and Colpophyllia natans (4% of total). Acropora cervicornis, a once-dominant fore-reef species, was rare. In the multidimensional scaling analysis based on mean species density in each assessment, outliers included Abaco (ID#1) and San Salvador (ID#3) in the Bahamas, the Flower Gardens (ID#5) andVeracruz, (ID#6) in the Gulf of Mexico, Costa Rica (ID# 19), and the Abrolhos (ID#20). Cluster analysis revealed two clusters of high similarity (70% Bray Curtis similarity), shown in gray on the MDS ordination in Figure 6A. Yucatan (1D#7) and Akumal/Xcalak (ID#8; also in the Yucatan) were in the first cluster; the second cluster consisted of Andros (ID#2), Turks and Caicos (ID#4), Maria la Gorda (1D#1 1), the Cayman Islands (ID# 12) and the Virgin Islands (ID# 13). Small (<2 cm) coral density averaged 4/m for the region with shallow reefs 9 9 having fewer of these “recruits” (3.3/m ) than deep reefs (4.4/m ) (Tables 3 A, 3B). By far the highest reported densities (9-15/m ) for both shallow and deep reefs were in the Abrolhos (ID#20). On deep reefs, low densities (<2/m ) were observed in the Flower Gardens (ID#5), Veracruz (ID#6), Akumal/Xcalak, (ID#8), and Maria la Gorda (ID#1 1) (Table 3B). Their species richness was also higher for deep (X^35 coral species) than for shallow (X^15 species) reefs, although this was partly a function of survey effort. Species composition of the recruits did not reflect the large (>25 cm) framework-builders present for the region as a whole. Brooders such as Porites spp. and Agaricia spp. dominated the assemblages with Porites astreoides being the most abundant species in both shallow and deep reefs. Broadcast spawners were far less common and the Montastraea annularis species complex only comprised ~2% of small corals observed overall. Acropora palmata (0.8% on shallow reefs) and /I. cervicornis (0.2% on deep reefs) were both rare. An exception was Siderastrea siderea, for which the overall percentage of small corals was similar to, or greater than, adult abundance (3% versus 3% in shallow and 12% versus 7% in deep for <2 cm and >25 cm colonies, respectively). Live cover and size. Live coral cover averaged 18% for shallow reefs with the highest (-38%) value reported for the reef crest sites on Andros (ID#2) and the lowest (-3%) for a patch reef in Costa Rica (ID# 19) (Table 3 A). On deep reefs, live coral cover ranged from 3% to 58%, averaging 26% for the region as a whole (Table 3B, Fig. 7A). The highest observed coral cover on deep reefs was in the Flower Gardens (ID#5). Four other assessments that are located in the southern-southeastern Caribbean [Los Roques (ID# 16), Bonaire (ID# 17), Cura9ao (ID# 18)) and St. Vincent (ID# 15)] also had coral covers that exceeded the regional average (Table 3B). Apart from Cura9ao (ID# 18), Abrolhos (ID#20), and Costa Rica (ID #19), the density of large (>25 cm) colonies generally correlated with live coral cover for the other 15 survey areas (r = 0.8, p<0.001, n = 15, deep depths only). Shallow reefs had overall lower mean densities (-7.9/1 Om) than deep reefs (-9.3/1 Om) (Fig. 7B), although a significant correlation (r=0.7, p<0.01, n=13) existed between the shallow and deep coral densities for areas in which both depths were assessed. The density of large (>one meter maximum diameter) colonies of 14 A stony coral species Dimension 1 B AGRRA fish species Figure 6. (A) MDS ordination of 20 survey areas (deep sites only) based mean species density for 20 most common large stony corals (=25 cm diameter). Groupings indicate Bray-Curtis similarity of 60% or more for all 20 surveys. (B) MDS ordination of 17 survey areas where fish were surveyed using belt transects (deep sites only). The clustering is based on mean species density for 40 most common species (excluding haemulids). See Table 1 for ID codes and text for discussion. 15 80 A Live stony coral cover % 60 40 * mean 20 r 0 SL i]“* _ * 1 ]S[ _ ^ * HR - d JiL I'D VO 200 C Max. diameter of the >25 cm Montastraea annularis complex E o 150 100 50 * * 71 cm 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 D Density of small (< 2 cm) stony corals * j -- mean A A He _ s n.iii ii _ _ * [ 1 Si: ii :] i 1 2 3 4 5 6 7 8 9 1011 1213141516171819 20 (D CO CD T- N- CM CO CSJ II II II tl II II C C C C C C. Tj- CM O O CD CM T- T- T- '•d- CM II II II II II II II c c c c c c c CM CO CD CD ■»- CM N- II ti II II II II II C C C C C C C Bahamas Gulf of Mexico Western Central Eastern Southern Brazil Caribbean Caribbean Caribbean Caribbean Figure 7. Comparison between means and standard deviation of deep sites for all 20 AGRRA assessments. (A) live coral cover; (B) coral density (number of > 25 cm diameter corals/1 Om); (C) colony diameter of the Montastraea annularis complex, and (D) small (<2 cm) coral density. The stars indicate a parameter that was not measured and the dashed line indicates the mean for all surveys. See Table 1 for ID codes. 16 Montastraea spp. at deep sites ranged from 0.01 to 3.1/10 m. Areas showing greatest densities were Flower Gardens, Bonaire, Los Roques, Cura9ao, while areas with lowest densities were Brazil, Veracruz, Abaco, and Costa Rica. Mean size (based on maximum colony diameter) for 17 of the 18 most common species of corals was generally larger in shallow sites (n=53) than in deep sites (n=249). The mean diameter for all 7,398 colonies of the Montastraea annularis complex of species (M annularis, M. faveolata and M. franksi) averaged 7 1 cm for deep reefs in the region (Fig. 7C) with the largest colonies having been observed in Abaco (ID#1), Los Roques (ID# 16) and the Flower Gardens (ID#5). Deep sites with higher than average live coral cover and abundance generally had larger than average colony diameters for the Montastraea annularis species complex (especially Los Roques and the Flower Gardens). Exceptions included Abaco and Akumal/Xcalak (ID#8) which had above average sizes for this species complex and lower than average coral cover (about 13% and 20%, respectively), and Bonaire (ID# 17) where coral cover was the second-highest recorded (~46%) but maximum colony diameter for the M. annularis complex was only “average” (Fig. 7A,C). Size frequency distributions for five of the numerically-most-common taxa, Montastraea annularis, M. faveolata, M. cavernosa, Siderastrea siderea, and Diploria strigosa (all broadcast spawners; two being included in the above analysis), showed that most colonies were in the 30-40 cm size class (Fig. 8). Acropora palmata (another broadcaster) was most frequently observed in the 100-120 cm-size class. Brooding species (e.g., Agaricia spp. and Porites spp.) were predominantly distributed in smaller size classes. Overall, coral sizes were fairly similar across the region, with greatest variability evident on smaller spatial scales, particularly within sites. Areas thought to have more marginal conditions for coral growth showed consistently larger coral sizes of certain species as follows: Veracruz (S. siderea, M. annularis)', Abaco (M. annularis, Colpophyllia natans)', and Costa Rica (D. strigosa). Mortality. Recent partial mortality for the >25 cm colonies (expressed as the mean proportion (%) of the colony’ surface area as seen from above) ranged from 0.1 to 27% across the 20 survey locations with a regional average of 4% for both shallow and deep reef types. High levels of recent mortality (~>15%) were observed on deep reefs in Andros (ID#2) and Yucatan (ID#7) (Fig. 9A) while the highest levels on shallow reefs occurred in Akumal/Xcalak (ID#8). Low recent mortalities (~<1%) were observed in Veracruz (ID#6), Bonaire (ID# 17), and the Abrolhos (ID#20) for deep reefs (Fig. 9 A) and in Veracruz, San Salvador (ID#3), and Cayman (ID# 12) for shallow reefs. Recent mortality levels were not correlated between shallow and adjacent deep reefs within the same survey (Tables 3A, 3B). Old partial mortality (excluding the standing dead corals) averaged 22% for deep reefs and 27% for shallow reefs (Tables 3A. 3B). Deep reefs in the Flower Gardens (ID #5) and Veracruz (ID#6) had the lowest reported old mortality (~<10%) (Fig. 9B) while shallow reefs in the TCI (ID#4), Belize (ID 10), and the Virgin islands (ID# 13) had the highest old mortalities (~>40%). Standing dead corals (expressed as a percentage of the total number of colonies assessed) were significanty more common (p<0.05, n=10 assessments) on shallow reefs (14%) than on deep reefs (4%) largely because of the 17 % 40 20 A Recent mortality A. 'dblS J^tir dfi rS F ^ - 4% 60 40 B Old mortality % 20 i 0 |l_ , * A Oh * rnl& Aflft- ■T-y - ^22% 80 1 C Total mortality (including standing dead except ID #6) 60 J % 40 I 20 ^ 0 I fr' Jna, * * rfi ^ j-i- ^ 28% *to 80 D Montastraea annularis complex -[-Total mortality (including standing dead except ID #6) 60 J % 40 20 niti a 32% 1 2 34 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 II II II II II II II II II II II c c c Bahamas Gulf of Mexico Western Central Eastern Southern Caribbean Caribbean Caribbean Caribbean Brazil Figure 8. Comparison (as means and standard deviations) of the in deep (> 5 m) sites in all 20 AGRRA surveys: (A) recent partial colony mortality; (B) old partial colony mortality; (C) total partial colony mortality (including standing dead); (D) Montastraea annularis complex total partial colony mortality (including standing dead. The star indicates a parameter that was not measured and the dashed line indicates the mean for all surveys. See Table 1 for ID codes. 18 >. o c: (D 13 O’ o 40 ^ CO ■e o 20 E "to ■■C CO CL ;o 0 O >» o c 0 o O’ 0 IL. 40 30 20 10 0 tE Diploria strigosa !n= 1162 40 20 0 10 20 0 F Acropora palmata n= 993 20 o o o o o O o o o o OO 1 in 1 CO 1 1 00 I CD o T“ X — in o o o o o CD 1 1 CM OO in CD OO o CD o o A o o o o o o O O o Y in 00 CM CD o M- 00 CM CM 1 o i 1 CM 1 CM CM 1 OO 1 OO CM CD o o o CD o o A o OO CM CD o T- T— CM CM 00 CO •C o E CO ■■e CO Q. o Max. diameter size class (cm) Max. diameter size class (cm) Figure 9. Size frequency distributions for six commonly observed coral species created by combining data (shallow and deep) from all assessments: (A) Montastraea annularis; (B) Montastraea faveolata', (C) Montastraea cavernosa', (D) Siderastraea siderea', (E) Diploria strigosa', and (F) Acropora palmata. Also plotted for each size class is partial old colony mortality, with the scale indicted on the second Y axis. Note the significant increase in partial mortality with increasing size, particularly within the smaller size classes, in each of these species. 19 abundance of dead colonies of Acropora palmata. Areas in which many of the shallow colonies were standing dead included Belize (ID# 10) and San Salvador (ID#3) with 17% and 25% respectively, and Los Roques (ID# 16) with 52% (Tables 3 A, 3B). Total mortality (including colonies that were standing dead) averaged 28% of the colony surfaces in deep reefs (Fig. 9C), and 39% in the shallow reefs Mortality levels (recent and old) varied dramatically among the individual coral taxa (Table 4). Agaricio tenuifolia had the highest recent partial mortality (>16% of colony surfaces affected) followed by Montastraea franksi (8%). Agaricia tenuifolia also had the highest old partial mortality (38%) followed by M. annularis (33%) (Table 4). Species with the largest number of standing dead colonies were Acropora palmata (32%), Agaricia tenuifolia (29%), dead Acropora cervicornis (18%). On average, these three species had three times or more the amount of standing dead corals compared to the other species in which standing dead was recorded. Total mortality values in the M. annularis complex on the deep reefs (Fig. 9D) were greatest in Abaco (ID#1) and San Salvador (ID#3) and lowest in Costa Rica (ID# 19). For many of the taxa, smaller corals (~30 cm diameter) showed either no mortality or 100% mortality (i.e., they were completely dead), whereas larger corals (>60 cm diameter) most commonly displayed partial-colony mortality. Recent mortality showed no significant correlation with colony diameter; however, old mortality levels generally increased with increasing colony size through the smaller size classes. Thus increases in old mortality were most evident in the lower size classes up to the mode of the population, above which fluctuations were more apparent (Figure 8). Disease, bleaching, overgrowth, and predation. Approximately 6% of the surveyed corals from all sites combined were infected with an identifiable disease. Signs of disease were observed in at least 20 coral species and, as illustrated in Table 4, many diseases were not species-specific but affected a variety of species. Stephanocoenia intersepta and Acropora cervicornis had the highest prevalence (21 and 13%, respectively) while the spatially dominant M annularis species complex also had high occurrence of disease (~10%). The most commonly observed disease was white plague, followed by black-band, yellow-blotch (yellow-band), white-band, dark spots, and white pox (patchy necrosis) (Table 4). Black-band was most commonly observed in M. franksi, white plague in M.faveolata and M. annularis, white-band in Acropora cervicornis and A. palmata, and yellow-blotch in M. annularis. In some cases, the disease type was not identified or specified for several species with apparently high prevalences of disease (e.g., A. cervicornis, S. intersepta, A. palmata). Areas with the highest prevalence of disease included Andros (ID#2) and Akumal/Xcalak (ID#7), where 18% and 14%, respectively, of all deep (>5m) colonies were infected (Table 3B). These areas were assessed during August, 1998 and March, 1999. Prevalences of yellow-blotch were relatively more common in the southern Caribbean (e.g., in Cura9ao, Bonaire, and Los Roques). An average of 6% of the surveyed corals from all sites combined showed signs of bleaching, most often occurring as “partly bleached” or “pale”. Four areas reported greater than 20% of colonies bleached (ID#sl0, 13, 14 and 17), while only four areas had less than 2% bleached (ID#s 4, 12, 18 and 20). Differences in the amount of bleaching 20 and disease varied among taxa (Table 4). Montastraea annularis species complex (13- 15% of colonies) and Millepora complanata (14%) exhibited the highest prevalence of bleaching. Bleaching was also often observed in Diploria spp., Siderastrea siderea, and Acropora palmata. Macroalgal overgrowth was reported on an average of 4% of the colonies assessed for the entire region. Unusually high occurrences (>30% of all colonies) of macroalgal overgrowth were recorded in Lighthouse Atoll, Belize (ID#9) and Costa Rica (ID# 19). However in 9 of the 1 7 deep surveys for which overgrowth was noted, fewer than 2% of the corals were affected. The percentage of colonies with evidence of tissue mortality due to fish bites was highly variable, being relatively high (>10%) in Costa Rica, Lighthouse Atoll and the Virgin Islands (ID# 13). Recent mortality caused by predation by snails (primarily CoralUophila abbreviata) or worms {Hermodice carunculata) was highest in shallow sites (e.g., Akumal/Xcalak (ID#8), Virgin Islands and Los Roques (ID# 16)). Occurrences of damselfish algal gardens (as the percentage of affected colonies) were also highly variable, in part because of inconsistent reporting, but were recorded on an overall average of 6% of the corals. The highest occurrences (>40%) were at Lighthouse Atoll and Maria La Gorda (ID #11), while Costa Rica and the Turks and Caicos (ID #4) were the only assessments for which no damselfish algal gardens were reported (Tables 3A, 3B). Algae and Diadema Algal communities on deep reefs were composed, on average, of turf algae (48% relative abundance) with crustose coralline algae (29%) and macroalgae algae (23%) being relatively less abundant. Deep reefs with highest values for turf algae (>70%) included the Flower Gardens (ID#5) and the Abrolhos (ID#20) whereas those with the lowest (<21%) included Andros (ID#2), Cayman (ID# 12) and St. Vincent (ID# 15). Elevated (>35%) macroalgal relative abundances were found at three areas in the Bahamas (ID#1, 2, 3) and several areas in the central and eastern Caribbean (ID#sl 1-13, 15) (Fig. lOA). Macroalgal relative abundance was low (~5 %) in the Flower Gardens, windward Netherlands Antilles (ID# 14) and Abrolhos and rare (<1 %) in Los Roques (ID# 16). Shallow reefs (< 5 m) had similar algal compositions (turfs>crustoses> macroalgae), although between-site variability was considerably higher. Macroalgal canopy heights were similar in deep (2.2 cm) and shallow (2.3 cm) reefs when averaged for the region. On deep reefs (Fig. lOB), the highest (>4 cm) canopy heights were recorded in Costa Rica (ID# 19) and the Yucatan (ID#7), and the lowest (<1 cm) were in the Flower Gardens (ID#5) and Bonaire (ID# 17). When macroalgal canopy height and relative abundance are considered together as a macroalgal index, there was a range of nearly 20-fold (Fig. IOC) between the lowest indices (<12) in Los Roques (ID# 16) and the highest index (>200) in Andros (ID#2). The range of values for the relative abundance of crustose coralline algae was moderate (~30%) for deep sites and in no cases exceeded 50% (Fig lOD). The macroalgal xrustose coralline ratios (not shown) averaged 1.1 (range 0.1 - 3.5) with Bonaire, Cura9ao (ID# 18), Los Roques, TCI (ID# 4), and the windward Netherlands Antilles (ID# 14) all having relatively low (<0.6) ratios. A weak, non-significant 21 Figure 10. Comparison (as means and standard deviations) of algae in deep (> 5 m) sites in the 20 AGRRA areas: (A) relative abundance of macroalgae; (B) macroalgal canopy height; (C) macroalgal index, and (D) relative abundance of crustose coralline algae. The star indicates a parameter that was not measured and the dashed line indicates the mean for all surveys. See Table 1 for ID codes. 22 relationship was found between the percent live stony coral cover and the relative abundance of macroalgae (p=0.17, r^=0.12). Statistically significant relationships were found between canopy height and both live coral cover (p<0.01, r =0.37) and large coral density (p<0.01, r^=0.38) (Fig. 1 1). Canopy heights were also positively related to the density of Siderastrea siderea (p= 0.015, r =0.33) and negatively related to the density of Colpophyllia natans (p=0.03, r^=0.29). Densities of Diadema antillarum, both within and among surveys, were highly variable, with no individuals reported in about half (8/17) of the areas for which data are available (Table 1). Mean urchin density for all surveys reported in this volume was 2.9/100 m with the highest densities (23/100 m ) occurring in Costa Rica. Fishes A total of 71,102 fishes were sighted in belt transects consisting of 49,888 individuals in the deep sites and 21,214 in the shallow sites (Fig. 12). Of these, a total of 10,073 (14%) were surveyed in Brazil (See Kikuchi et al., this volume). For the non- Brazil sites, the most abundant of the AGRRA families were acanthurids and scarids, followed by haemulids, lutjanids, pomacanthids, balistids, and serranids. The results of the multidimensional scaling (MDS) ordination based on mean AGRRA species abundance recorded at deep sites for the 17 belt-transect assessments is shown in Figure 6B. Areas with the least similarity in fish community structure [Abrolhos (ID#20), followed by Costa Rica (ID# 19), Abaco (ID#1), Los Roques (ID# 16), St. Vincent (ID# 15), and the Flower Gardens (ID#5)] are mostly characterized by unusual environmental conditions. Cluster analysis revealed three groups, each with strong similarity (Bray-Curtis similarity of greater than 60%, areas shaded gray in Figure 6B), corresponding to areas in the southern Caribbean, parts of the Bahamas and the western and central Caribbean, and the central-eastern Caribbean regions, respectively. Total AGRRA fish densities were nearly twice as high in shallow (85/100 m ) as in deep (49/100 m ) sites, mainly as a result of higher densities of haemulids and acanthurids in <5 m. For deep sites, comparisons among the 17 areas with comparable data records indicate that fish densities were highest in the Abrolhos (ID#20), followed by Los Roques (ID# 16) and Cura9ao (ID# 18), while the lowest values were found off Andros (ID#2) (Fig. 13A). Herbivores dominated the deep fish assemblages (average of 35/100 m ) at approximately six times the density of carnivores (average of 6/100 m ) (Fig. 13B,C). Excluding data from the Abrolhos, total fish biomass (not shown) displayed a five-fold difference among surveys, ranging from 2,600 g/100 m in Akumal/Xcalak (ID#8) to 12,640 g/100 m^ in Los Roques. Using the Bryant et al (1998) Ree/s at Risk overexploitation threat classification (see Table 1), total AGRRA fish biomass was 6,459 g/100 m for the “high exploitation” reefs, 5,943 g/100 m for reefs under “intermediate exploitation”, and 6,846 g/100 m^ for reefs experiencing “low exploitation”. No significant difference in total fish biomass could be detected among the three different exploitation-threat categories using a 1-way ANOVA analysis. Herbivores. Deep reefs in Abaco (ID#1), the Abrolhos (ID#20), and Los Roques (ID# 16) all had higher than average densities of herbivorous fishes, whereas lower than 23 A B Figure 11. Regression plots of fish, coral, and algal indicators: (A) mean herbivore density (acanthurids, scarids >5 cm and Microspathodon chrysurus) and mean macroalgal index for 17 assessments; (B) algal canopy height and coral density (stony corals >25cm diameter/per transect). 24 Figure 12. Composition of AGRRA fishes for major families for all assessments except Brazil (ID #20) combined for (A) shallow sites (<5 m), and (B) deep sites (> 5m). Includes only fishes sampled within a maximum of 10 transects per site. 25 Figure 13. Comparison between means and standard deviation of deep sites (>5 m) for 17 AGRRA assessments: (A) all AGRRA fishes, (B) herbivorous fishes (acanthurids, scarids >5 cm and Microspathodon chiysurus); (C) carnivorous fishes (serranids, lutjanids, haemulids >5 cm). The star indicates a parameter that was not measured and the dashed line indicates the mean for all surveys. 26 average densities were found in Andros (ID#2), the Turks and Caicos (ID#4), Yucatan (ID#7), and Maria la Gorda (ID#1 1) (Fig. 13B). Herbivore biomass (not shown) ranged from a high of 7,484 g/100 m^ in Costa Rica (ID# 19) to a low of 1,508 g/100 m^ in the Turks and Caicos). Microspathodon chrysurus (yellowtail damselfish) had a mean density of 2.2/100 m with greatest numbers (>4/100 m ) reported in Los Roques, Lighthouse Atoll (ID#9), and Costa Rica and the lowest (-0.1/100 m^) in Andros. The density of acanthurids for all surveys combined averaged 1 1.8/100 m . Acanthurids were widespread throughout the region and all three species, Acanthurus coeruleus (blue tang), A. chirurgus (doctor fish) and A. bahianus (ocean surgeon), were present in all 17 areas surveyed. A. coeruleus was most abundant (averaging 5.7/100 m ), with a high of 30.7/100 m^ in Bonaire (ID# 17), followed by bahianus (average = 5.3/100 m^), which reached a high of 8.7/100 m in the Virgin Islands. Acanthurus chirurgus was least abundant (average of 1.0/100 m ) except for the Abrolhos where its density (13.7/100 m ) was unusually high. Scarids had a mean density of 13.7/100 m and were most abundant in the eastern and southern Caribbean. The highest scarid densities (36.0/100 m ), which occurred in Los Roques, were nearly five times those reported for Maria La Gorda) and Lighthouse Atoll (Fig. 14A). Parrotfish species composition was similar across the region except for the Abrolhos which contained a Brazilian endemic, Scarus trispinosus (greenlip parrotfish), not found elsewhere in the western Atlantic (Fig. 14A). Scarus croicensis (striped), Sparisoma aurofrenatum (redband), Scarus taeniopterus (princess), Sparisoma viride (stoplight) and Scarus vetula (queen) were the five most abundant parrotfish species overall, with mean densities of 3.8/100 m^, 3.6/100 m^, 3.1/100 m^, 2.8/100 m^, •y and 1.1/100 m , respectively. Large-sized parrotfishes, including Scarus guacamaia (rainbow), Scarus coelestinus (midnight), and Scarus coeruleus (blue), were observed only occasionally and were more common in the southern Caribbean than in other subregions. Regression analysis showed no significant relationship (p>0.05, n=17) between macroalgal index and herbivore density at the scale of the entire region (Fig. 1 1 A). Similarly, no significant relationships were found between herbivore density and any other algal indicator (e.g., macroalgal canopy height or relative abundance of macroalgae, turfs or crustose corallines). Regionally, total herbivore density or biomass showed no relationship with coral density or with such measures of habitat complexity as coral diameter or height, although significant relationships were found between individual scarid and acanthurid species. For example, densities of S. vetula (queen) and S. taeniopterus (princess) parrotfishes were positively related to live coral cover (r = 0.8, p<0.05, n==17; r = 0.6, p<0.05, n=17 respectively). The density of A. chirurgus (doctor fish) was negatively correlated to old partial mortality (r =-0.53, p<0.05, n=17) while that of A. bahianus (ocean surgeon) was negatively related to large coral density (r =-0.49, p<0.05, n=17). Carnivores. The density of carnivorous fishes averaged 6.0/100 m for the region and was similar in shallow and deep reefs. Lutjanids predominated the carnivore assemblages overall with nearly twice the densities of serranids. Proportionally more serranids occurred in deep reefs than in shallow, although the reverse was true for 27 0 Scarus trispinosus □ Sparisoma spp. ^ Scarus spp. □ Scarus guacamaia Scarus coelestinus ■ Sparisoma chrysopterum □ Sparisoma rubripinne B Scarus vetula ■ Sparisoma viride □ Scarus taeniopterus m Scarus croicensis □ Sparisoma aurofrenatum CM E o o E Epinephelus marginatus d Mycteroperca acutirostris m Mycteroperca interstitialis s Mycteroperca venenosa ■ Mycteroperca bonaci ® Epinephelus striatus Q Epinephelus adscensionis □ Mycteroperca tigris □ Epinephelus fulvus E3 Epinephelus guttatus * Epinephelus cruentatus Bahamas Gulf of Western Central Eastern Southern Brazil Mexico Caribbean Caribbean Caribbean Caribbean Figure 14. Stacked bar plot of fish family composition by survey (deep sites only) (A) scarid and (B) serranids. The star indicates a parameter that was not measured. 28 lutjanids. When comparisons are limited to the deep reefs (Fig. 13C), the areas with highest carnivore densities were the Abrolhos (ID#20) and Los Roques (ID# 16), while the lowest were found in St. Vincent (ID# 15) and Maria La Gorda (ID#1 1). Excluding the Abrolhos, carnivore biomass displayed a 40-fold difference across all other assessments, ranging from a low of 191 g/100 m^ in Akumal/Xcalak (ID#8) to a high of 4,515 g/100 m^ in Los Roques (Fig. 13D). No significant difference (p = 0.84, n = 3) was detected in total carnivore biomass between highly threatened reefs (~ 1,1 00 g/100 m^ ) and either medium- (1,204 g/100 m^) or low-threat reefs (-1,280 g/100 m^). The most abundant lutjanids were Ocyurus chrysurus (yellowtail snapper), Lutjanus apodus (schoolmaster snapper), and L. mahogoni (mahogany snapper) at densities of 1.6/100 m , 0.8/100 m , and 0.4/100 m , respectively. Densities of Ocyurus chrysurus recorded in the Abrolhos (ID#20) were one of the highest (-23.1/100 m^) for any species. Serranids were present in all surveys at densities of less than 2/100 m and were most abundant in the windward Netherlands Antilles (Fig. 14B). In all surveys, except Andros (ID#2) and the Abrolhos, smaller-bodied species including Epinephelus cruentatus (graysby), E. guttatus (red hind), E. adscensionis (rock hind), and E. fulvus (coney) dominated the grouper assemblages. The density of coneys decreased substantially in surveys conducted in the southern Caribbean where graysbys were proportionally more common (Fig. 14B). The most commonly seen large-bodied grouper was Mycteroperca tigris (tiger) with a mean abundance of 0.4/100 m . Epinephelus striatus (Nassau grouper) was rarely seen except in the Bahamas subregion with Andros having the highest density (0.2/100 m ). Single [E. marginatus (dusky), Mycteroperca rubra (comb)] to no [Epinephelus itajara (goliath), E.morio (red), and M. phenax (scamp)] sightings were recorded in transects for a number of serranids on the AGRRA list. Sphyraena barracuda (great barracuda) was recorded in 10 of the 17 surveys, although usually at very low numbers (<0.2/100 m ). Index of Reef Health To evaluate the overall condition of each survey on a relative scale, a preliminary biotic reef health index was developed for the deep (> 5 m) dataset utilizing 1 3 of the AGRRA indicators: live coral cover, large (>25 cm) coral density, small coral density (Fig. 7D), maximum diameter of the Montastraea annularis complex , recent partial mortality, old partial mortality, total mortality (including standing dead), prevalence of coral diseases, macroalgal index, relative abundance of crustose coralline algae, Diadema density, herbivorous fish density, and carnivorous fish density. Only 1 7 areas could be analyzed because of missing [Veracruz (ID#6)] or incomparable [(San Salvador (ID#3) and Belize (ID# 10), where fish were quantified in cylinders] fish data. A similarity matrix was calculated and analyzed using a Bray-Curtis similarity cluster dendrogram (Fig. 15). Three broad classes, distinguishable at the 70% similarity level, were labeled as “worse,” “average,” and “better” health categories. Only four areas fell into the “better” grouping: the Flower Gardens (ID#5), windward Netherland Antilles (ID# 14), Los Roques (ID# 16), and Bonaire (ID# 17), each of which was characterized by low recent mortality, low prevalence of diseases, low macroalgal index, and relatively high fish densities. Seven areas fell into the “average” grouping and most of these had the majority of their 29 40 X 0 •D C 60 1 (D CO ■•c 13 o I >» CO L_ CD 80 100 “Worse” “Average” “Better” Figure 15. Bray Curtis similarity diagram of proposed reef health classification based on 13 AGRRA indicators. See Table 1 for ID codes and text for discussion. 30 indicators near the regional norms. Surveys in the most unfavorable class (“worse”) were Abaco (ID#1), Andros (ID#2), Yucatan (ID#7), Akumal/Xcalak (ID#8), Maria La Gorda (ID#1 1), and Costa Rica (ID# 19). These areas were generally typified as having high recent mortality, high prevalence of disease, moderate-to-high macroalgal index, and low fish densities. DISCUSSION Sampling Issues Sampling effort. Differences in the sampling effort carried out for each AGRRA assessment resulted in part because of differences in the areal extent each team attempted to characterize as well as logistical sampling constraints (e.g., time, support, and accessibility). The fewest sites (two) were for the Flower Garden Banks (ID#5), which was also the deepest (~20 m) and furtherest offshore of the reef areas studied. The greatest effort was for the Cayman Islands (ID# 12) where a total of 42 sites were assessed. Determining how much effort to allocate towards characterizing sites depends on the indicator and on its variance at different spatial scales. Results of the power analyses at the site scale suggest that, for a number of AGRRA benthic indicators, the greatest gain in terms of precision versus effort occurs within the first six transects which corresponds to roughly 60 corals and 40 quadrats. For the more specific indicators of rare species (i.e. abundance of large-bodied groupers, size frequency distribution of some corals), higher overall sampling would be necessary to get an adequate sample size to test significant differences. The coefficients of variation (CV), defined as the standard deviation divided by the mean), for 1 2 of the AGRRA indicators at four spatial scales (deep sites only) are given in Table 5. For most indicators, the highest CV occurs at the smallest spatial scale (<0.1 km), however, there is still significant variation at the area and subregional scales (~1-100 km). This inverse relationship between CV and spatial scale suggests that in terms of sampling effort either slightly more effort should be allocated at the site scale (i.e., number of transects sampled per site) in comparison to the larger scales (i.e., number of sites sampled per area and number of areas sampled per subregion) and/or to observer training (see below). It also indicates that the size of the regional “signal” for some indicators (e.g., partial old mortality; herbivore fish abundance) is quite small but comparatively large for others (partial recent mortality, macroalgal canopy height, carnivore abundance). Site selection and habitat classification. Considerable variance in AGRRA indicators at all but the smallest spatial scales can arise by sampling slightly different reef types. In this initial synthesis, only two categories of reefs were recognized (deep and shallow) and all sites in the deep category were used because the purpose was to compare the assessments presented in the volume. A majority of the deep sites were located on fore-reef slopes but some were located in other habitats. Habitat variation clearly has an influence on some AGRRA indicators (e.g., coral community composition, standing dead 31 coral) and these sites should be factored out in future comparisons. Reef slopes themselves can also have different morphologies that influence fish densities and other AGRRA parameters. For example, fore-reef slopes around the Turks and Caicos (ID#4) and on the leeward sides of Los Roques (ID# 16) and Bonaire (ID# 17) are characterized by sharp drops (walls) at depths of approximately 10m (the depth at whieh many surveys were undertaken). These “edges” tend to attract fishes at higher abundances compared to adjacent zones both shallower and deeper. Site selection and classification are even more critical for shallow-reefs where large differences in community coral composition, fish abundance, and algal parameters can occur across very small spatial scales (<100m). I have observed that reef habitats surrounded by areas of deep water (or similar sites with high relief surrounded by vast areas of no relief) will tend to attract fishes (“oasis effecf ’) and have higher fish densities than areas wherein habitat is more evenly distributed. In the Abrolhos (ID#20), the unusually high fish densities are attributed, in part, to the shallowness and small size of the reefs. As the number of AGRRA assessments in the database grows, geomorphic and structural complexity characteristics should also be used to help classify reef types and design more representative sampling strategies for field assessments. Methodology and observer bias. Observer bias is a potential contributor to variance in the benthic and fish data and needs to be addressed in any study that relies on more than a single observer to collect data. Many of the participants contributing to this synthesis had eonsiderable experience with Caribbean reefs and, with few exceptions, one or more of the observers in each team had received formal training in the AGRRA protocols. Based on our workshop experienees, variance among observers who have undergone consistency training is small compared with the total variance at all spatial scales for any given indicator. When observers have not undergone such training, a number of the AGRRA indieators can be scored systematieally in ways that lead to bias in the data. This is particularly true for fish where small differences in belt-transeet width, length and swim time can lead to systematie differences in fish density estimates (Sale, 1998). Several of the surveys in this volume used different transect sizes or even different methods. Given the costs and time to collect AGRRA data, particularly from remote locations, it is recommended that the protocol be followed more closely in future field efforts and that more effort be spent on minimizing observer bias through consistency training of all observers with eore AGRRA personnel. Stony Corals Community structure. A mean live stony coral cover of 26% for deep sites is considered fairly representative of the region during 1998-2000, although perhaps somewhat elevated sinee the most structurally complex sites were sometimes selectively targeted by AGRRA assessors. Historic baselines for eoral cover do not exist in most areas, but a pre-1980 regional estimate based on available coral cover data was summarized by Gardner (2002) at between 40 and 50%. Thus Western Atlantie reefs have undergone signifieant losses over the past several deeades yet despite many large-scale disturbances (e.g., the 1983-84 die-off of Diadema antillarum, numerous outbreaks of 32 disease and bleaching events) a fair amount of living coral remains. In some areas, however, and not reflected in these percentages, the original large corals that broadcast larvae have likely been replaced by smaller brooding species. More substantial changes have probably occurred in the shallow (<5 m) reef crests and patch reefs where faster-growing acroporid corals once dominated (e.g., Goreau, 1959; Geister, 1977; Adey, 1978; Gladfelter, 1982). Results of the AGRRA assessments for the shallow reefs presented in this volume indicate an overall live coral cover of 18%; however, the small and spatially disproportionate sampling and high variability in reef types that were assessed preclude using this average as representative of the region. AGRRA assessments since 2000 have been directed specifically to more systematically sample shallow reefs. For shallow and deep reef sites, the composition indicates typical species dominance patterns on western Atlantic reefs and reflects the amount of survey effort for a particular reef type. Coral species composition and abundance data for the 20 areas are influenced by: (1) environmental and/or biogeographic factors; (2) reef types that were assessed; and (3) their disturbance histories. One of the two clusters in the MDS analysis of the deep reefs is caused in part by the geographic location of its areas (ID#s 2, 4, 11, 12, 13) in the insular Caribbean (Fig. 6A). Outliers represent geographic extremes or areas that are currently marginal for reef growth. For example: Abaco (ID#1), the furthest north, is exposed to large Atlantic swells and to cooler water temperatures; Abrolhos (ID#20), the furthest south, is in a different biogeographical subregion and also experiences open ocean swells; both Costa Rica (ID# 19) and Veracruz (ID#6) are heavily influenced by sediment runoff; and the Flower Gardens (ID#5), the deepest of the assayed reefs and somewhat isolated from other reef systems in the northwestern Gulf of Mexico, lack several common Caribbean scleractinian species (Roberts et ah, 2002). The distinctiveness of the deep reefs in San Salvador (ID#3) (compared to other deep reefs) is thought to result from the way sites were grouped using the 5m cut off depth, which resulted in four “other” reef types being grouped with three reef-slope sites. The abundance of “recruits” (estimated as small corals with diameters of <2 cm) is an important indication of a reefs potential for growth and for recovery after major disturbances. The total area sampled at each site (~2-3 m ) was probably too small for these data to be highly reliable at the site scale. Patterns at the subregional level were not evident, but larger scale comparisons indicate that corals are currently recruiting most successfully in the Abrollhos (ID#20) and on the deep reefs in Andros (ID#2) and the Virgin Islands (ID# 13). That the AGRRA data shows species composition of coral recruits often did not reflect the major coral-reef builders present is similar to other studies (e.g., Rogers et al., 1986). Mortality. The AGRRA data support previous findings that show partial mortality differs among coral species and partially reflects the life history strategies and population dynamics of species and their susceptibility to various physical and biological factors (Bak and Luckhurst, 1980; Hughes and Jackson, 1980, 1985; Meesters et al., 1997). Large, long-lived broadcasting corals such as Montastraea spp. tend to exhibit higher amounts of partial tissue mortality while smaller, short-lived brooding species (e.g., some species of Porites and Mycetophyllia) tend to exhibit either complete mortality or no 33 mortality. For nearly all species, levels of old partial mortality increase with colony size (particularly in the early size classes) suggesting that corals are more likely to suffer severe, irreversible tissue damage the longer they live (e.g., Ginsburg et al., 2001). Since between-reef variability of mortality is strongly influenced by the species composition and sizes of the corals present in each reef (Bythell et al., 1993; Meesters et al., 1996), these variables should be considered when examining patterns at diverse spatial scales. The mortality signal is also a function of how long the corals remain within the “recent” state before transitioning to the “old” state and finally to unidentifiable substrate or rubble. This rate of transformation is strongly influenced by sedimentation, bioerosion, and overgrowth all of which vary both spatially and temporally. Given these influences and geographic variations in the prevalence of disease and bleaching events, hurricanes and other disturbances in 1998-2000, it is not surprising that levels of recent partial mortality (along with the amount of standing dead coral) were highly variable at nearly all examined spatial scales (0.1-1000 km). In contrast, old partial mortality showed much lower variation at larger spatial scales (>~0. 1 km) suggesting similar processes were affecting this indicator for the entire region. A useful distinction can be made between background mortality of stony corals caused by chronic stressors and acute mortality caused by intermittent major disturbances (see also Steneck and Lang, this volume). Some inferences can be made about each of these sources given the range of environmental conditions that affected the reefs sampled in this volume. Chronic or background coral tissue damage and mortality can result from a variety of low-intensity stressors including predation, disease, bioerosion, sedimentation, competition and abiotic perturbations (Meesters et al., 1996, 1997). Many instances of partial mortality are small and within the abilities of corals to regenerate new tissue; however, above certain sizes (~10cm ), dead skeletal areas tend to become overgrown or eroded by other organisms (Meesters et al., 1996; Clark, 1997). Under increasingly severe chronic conditions, especially for species with slow rates of regeneration (e.g., most massive corals), there will be increases in partial mortality followed by decreases in size structure, increases in the number of physiologically distinct colonies, and finally shifts in the species assemblages. Abaco (ID#1) and Costa Rica (ID# 19) are two of the most chronically stressed reef areas assessed in this volume. The deep Abaco reefs had the highest levels of total mortality for the M. annularis complex (Fig. 9) coupled with the lowest densities of large (>lm) colonies (Table 3B). Costa Rica had proportionally more colonies of Siderastraea, which is well-known for its high tolerance of sediment stress (Cortes and Risk, 1985) than any of the other 19 areas, but relatively low levels of total partial mortality for the M. annularis complex and of partial old mortality for all stony corals. These patterns underscore that differences in partial mortality are influenced by the species composition of corals present and coral size and should be considered together when examining patterns at various spatial scales. Establishing a present-day baseline for background coral mortality allows us to distinguish the magnitude of mortality following severe disturbance events such as disease epizootics, bleaching, or hurricanes. Based on the AGRRA data presented in this volume, our current best estimate of background (chronic) partial mortality on reefs can be approximated from the regional averages of <4% for recent mortality and <22% for old mortality. However, a more accurate estimate of chronic mortality is achieved by 34 removing those areas known to have been recently impacted by severe disturbance events (e.g., ID#sl,2,3,7,8,10). This results in an estimate of <2% for recent mortality and <21% for old mortality. Associating a time frame with these recent and old mortality signals would allow us to make an inference about the rates of mortality. Assuming the recent morality signature (i.e., raised calices around corallite, unless calices have been removed by parrotfishes) remains visible for approximately one year, present-day rates of annual turnover (“cumulative partial mortality”) might be on the order of 2%, excluding any regeneration of soft tissues. By extension, this would imply that old mortality signatures remain visible for approximately a decade, after which coral skeletons are no longer identifiable. Although these values appear to represent a reasonable upper limit of background mortality to expect on reefs today, it is unknown how this estimate compares to historic background levels when chronic stresses may have been quite different. Hughes and Connell (1999) argue that rates of coral mortality can be naturally quite high even in the absence of major disturbance events, thus historic mortality levels may not be drastically different from those observed today. If so, this would imply that 30-50 years ago a regional average of 2% recent and ~20% old mortality may have existed, and that decreases in live coral cover may be more a function of changes in the rates of coral regeneration and recruitment rather than of changes in the rates of coral mortality. Coral tissue damage that results in large lesions or even complete colony mortality often results from major disturbance events (e.g., disease eipzooitics, bleaching, hurricanes). Shortly after the disturbance, this type of acute mortality is reflected mainly in high values for recent mortality and standing dead. For example, high levels of recent mortality were related to the presence of certain coral diseases at the subregional and local scales (Table 4B) including black-band disease (Andros, Kramer et ah, this volume), white plague (Akumal/Xcalak, Steneck and Lang, this volume) and yellow- blotch disease (Cura9ao, Bruckner and Bruckner, this volume). Mass mortality events often affect corals regardless of size but certain species are more susceptible than others. For this reason, species-specific levels of recent mortality are probably a more reliable way of examining spatial and temporal variation in disturbance events. The amount of standing dead coral, often overlooked in current monitoring methods, is an excellent indicator for hindcasting past disturbance events possibly exceeding a decade (provided no major hurricanes have directly affected the area in the meantime and if rates of overgrowth and bioerosion are not too high). Reliablity in counting and identifying standing dead massive corals to the genus or species level depends on how much time has passed since the disturbance event and how much effort an observer puts into exposing and identifying their skeletons. The highest occurrences of standing dead coral on deep reefs were reported from Belize (ID# 10) and San Salvador (ID#3), two areas strongly influenced by the 1998 ENSO (see two papers by Peckol et ah, this volume). As the reefs examined included some deeper patch-reef and back-reef communities, how much this signal represents the ENSO disturbance history versus differences in reef type from those in most of the other assessments is unclear. In general, the standing dead signature is more dramatic on shallow reefs where dead Acropora palmata are easily identified and can persist upright for decades, particularly if they become encrusted with crustose coralline algae which makes the skeletons less susceptible to bioerosion (personal observations). 35 Diseases and bleaching. AGRRA data indicate that diseases were present throughout most of the wider Caribbean region in 1998-2000 with very few areas exhibiting no occurrences, not even on reefs removed from close human influence. These data are consistent with numerous reports suggesting that the prevalence, extent and type of diseases in Caribbean stony corals continue to increase (Richardson, 1998; Green and Bruckner, 2000; Wheaton et al., 2001). An unexpected observation from the AGRRA surveys was the high prevalence of disease in the Montastraea annularis complex observed on some of the deep reefs, particularly given their relatively remote locations (e.g.. Cayman, Kievman et al, this volume; Andros, Kramer et al, this volume; Virgin Islands, Nemeth et al, this volume). Outbreaks of disease (or epizootics) nearly always resulted in high spatial variability at the community and individual level and differentially affected species. On the regional scale, all size classes of corals were affected by diseases. However, on some local scales, size effects were evident. For example, higher prevalences of yellow-blotch disease were correlated with larger (>0.5 m) coral sizes in Cura9ao (Bruckner and Bruckner, this volume). Spatial patterns of disease occurrence on Andros suggested that at least some diseases may be highly contagious or spread rapidly and easily. Some of the deep reefs with high prevalences of diseases, for example, of white plague and black- band off Andros (Kramer et al., this volume) and of yellow-blotch off Cura9ao (Bruckner and Bruckner, this volume), also had high coral densities and complex reef structures dominated by the M. annularis species complex and may be more susceptible to rapid spread of a disease because of the close proximity of colonies. However, no diseases were observed on the well-developed reefs in the Flower Gardens (Pattengill et al., this volume) (although noted to exist in low amounts outside of transects) suggesting that; (1) some regions may be less exposed to diseases; (2) other stressors such as bleaching make corals particularly susceptible to diseases that are present on most reefs; and/or (3) some coral genotypes may be more resistant to diseases that others. The AGRRA data suggest a strong linkage between bleaching-related mortality and infectious diseases which could be due to increased pathogen activity and/or sensitivity of corals to their effects during periods of elevated sea surface temperatures. Recent mortality in areas that had been affected by mass bleaching during the 1 998 ENSO event was closely associated with the presence of black-band disease in Andros (Kramer et al., this volume) and white plague in Andros (ibid.) and the Yucatan (Ruiz et al., this volume; Steneck and Lang, this volume). However, these relationships are somewhat confounded since not all disease outbreaks are tied to temperature and because the impact of a disease will vary based on how long it lasts and how fast it spreads. Increased water temperatures have been related to elevated prevalence of five coral diseases (bacterial bleaching, black-band disease, white plague, aspergillosis and dark spots disease) (reviewed by Harvell et al., 1999; Rosenberg and Ben-Haim, 2002; see also Porter, 2001). Clumped distributions observed in the AGGRA surveys mentioned previously, may be due both to the intrinsic intensity of the disease and local patterns of increased sea surface temperature similar to a post-bleaching-related outbreak of black- band disease in Florida (Kuta and Richardson, 1996), whereas “normal background diseases” observed at the local scale on many of the other AGRRA surveys may have a 36 more random distribution pattern similar to that reported by Edmunds (1991). It is still unclear if bleaching causes corals to be more susceptible to opportunistic pathogens, and/or if pathogens normally present exacerbate levels of bleaching and bleaching related mortality. The relationships among disease, bleaching and mortality, and the temporal and spatial scales in which these processes are operating, will influence whether the effects of these events are transient or lethal (e.g., Kramer and Kramer, 2002). Condition of keystone corals. The AGRRA data provide further information concerning the regional decline of Acropora (Aronson and Precht, 2001), which is particularly evident for A. palmata as standing dead signatures in several locations (e.g., Los Roques, Villamizar et al., this volume; Tobago Cays, Dechamps et al., this volume). Acropora palmata still predominates species composition on many shallow reefs in the region although the proportion of dead-to-living colonies varies dramatically on the local scale. Surprisingly large stands of live A. palmata were found off Andros, Bahamas (Kramer et al., this volume) and, more recently, on the southwestern coast of Cuba (AGRRA database, unpublished data). In contrast, Acropora cervicornis was present on many of the deep reefs that were assessed, but was very sparse with no thickets or haystacks even remotely resembling the stands reported from Jamaica (Goreau, 1959), Bonaire (van Duyl, 1982) and Florida (Dustan and Halas, 1987) as recently as the late 1970’s to early 1980’s. The fragile nature of branches of^. cervicornis allows it to be broken down to rubble soon after death; thus the likelihood of observing standing dead colonies is substantially lower than for other species. Acropora cervicornis rubble can persist for several decades (Shinn et al., in press) however, and was noted in several areas (e.g.. Cayman, Curagao). Significant gaps still exist in our understanding of the geographical distribution and condition of acroporids. For some areas (Bonaire, Cura9ao, Cayman, Maria la Gorda, Costa Rica) shallow reef crests are of limited spatial extent and thus were not sampled. Elsewhere (Yucatan, Belize, San Salvador, Turks and Caicos, and the Virgin Islands), the extensive reef crests were not assessed because of inaccessibility or high wave energy. Greater sampling of shallow reef-crest habitats within the wider Caribbean should be a priority since so little region-wide information is currently available. Nevertheless, the historic range does not appear to be reduced or lost in either acroporid species although occurrence and abundance data suggest range reductions have occurred at the local scale. The presence of acroporid recruits was higher than expected in some areas (e.g., Andros, Cayman) and the existence of a few localized healthy populations of Acropora palmata is encouraging. However, very few A. cervicornis recruits were observed (Virgin Islands, Nemeth, this volume; Netherlands Antilles, Klomp et al., this volume) and there were few signs of significant reestablishment of/recolonization by this species. Four species within the Montastraea family (Montastraea annularis, M. faveolata, M. franksi and M cavernosa) numerically accounted for about half (~50%) of the frame-building corals on the deep reefs assessed in this volume except for Abrolhos where the endemic Mussismilia braziliensis predominated (Kikuchi et al., this volume). The density (as numbers/10 m) and sizes of Montastraea colonies can be used to indicate the stability of environmental conditions (e.g. Done, 1995) as well as infer characteristics 37 of mortality, regeneration, and recruitment (e.g., Bak and Meesters, 1998). High densities of large (>~1 m) colonies are indicative of environmentally stable “old growth” conditions (e.g., the Flower Gardens, Bonaire, Los Roques, and Cura9ao). In contrast, the lowest densities are found in areas that are marginal for reef growth (e.g., Abaco, Costa Rica). Smaller colonies (<30 cm) are less likely to have partial mortality but more likely to experience complete mortality, and thus may be a more sensitive indicator of environmental growth conditions. In current versions of the AGRRA protocol all corals equal to, or greater than, 1 0 cm maximum diameter are now sampled. High recent mortality coupled with the disease outbreaks described above were observed in the Montastraea annularis species complex at several areas during 1 999- 2000. Because of their slow rates of tissue growth (Hubbard and Scaturo, 1985), colonies with large injuries from fast-spreading diseases (e.g., white plague) have a high likelihood of impaired skeletal growth (Hughes and Jackson, 1985) or diminished reproductive output (Szmant and Gasman, 1990) and some may not recover at all. Given the importance of the M annularis complex as the current and historic principal framebuilder of intermediate-depth reefs, these large, and perhaps unprecedented, disease and bleaching impacts are cause for great concern. Improving our understanding of its dynamics (reproduction, settlement, mortality) and exploring any cases (genotypic, geographic, habitat, depth-dependent) of resistance and/or resilience to these impacts as may occur naturally should help better direct management strategies towards their persistence and recovery. Fishes Community structure. Fish community structure at any given locality can be attributed to a wide range of factors including environmental conditions, habitat complexity, and fishing and management regulations. Other factors such as predator-prey interactions (Hixon, 1991), larval supply and recruitment (Cowen, 2000), history of disturbance (Syms and Jones, 2000), quality and intactness of adjacent habitats (e.g., Munday, 2002), and natural biogeographic variation within species may also contribute to the overall variance in fish community structure. Most of the AGRRA species are found throughout the western Atlantic with the exception of the Abrolhos. Some species are known to be much more common in certain areas than others. For example, Nassau grouper are much more prevalent in the Bahamas than in other parts of the Caribbean. However, when combined into a carnivore index, these species differences become less distinct because there may well be other carnivore species present to fill the niche occupied by one that is not present. Herbivore species composition and sizes were remarkably consistent across surveys. That herbivore densities were more variable, with consistently higher numbers in the eastern and southern Caribbean and lower numbers in the Bahamas, Gulf of Mexico, Maria la Gorda (ID#1 1) and western Caribbean (Fig. 13B), may be partially related to geomorphic factors discussed above. Several significant relationships (positive and negative) between habitat variables and the abundance of individual fish species were observed and need to be explored in more detail before conclusions can be drawn regarding their significance. However, that 38 coral size, density, and partial mortality were not directly related to overall fish density (or biomass) nor with herbivore or carnivore indicators at the scale of the entire region suggests that some tropical fish-habitat relationships identified at smaller scales (e.g., Luckhurst and Luckhurst, 1978; Lindquist and Gilligan, 1986; Nemeth et ah, this volume) do not hold at larger spatial scales. One possible reason for the lack of a relationship at the regional scale is that guilds of fishes are utilizing the best available habitat within each of the assessment areas even though the habitats may be quite different from one area to another. Habitat quality in itself may not be the limiting factor determining fish abundance, but other factors such as larval supply, recruitment, and trophic predator-prey relationships may be more important. Clearly definitions of habitat quality vary and may include other measures not included in this analysis such as percentage of A. palmata (shallow), percentage of M faveolata (deep), or independent measures of rugosity. Since 2000, the newer version of the AGRRA protocols incorporates an independent measure of substrate rugosity which, in the future, will help to classify reefs based on relief and to better test relationships between structural complexity and other benthic, algal, and fish variables. Overfishing. Differences in fish community structure that can be either directly or indirectly linked to human fishing have been documented in numerous studies (e.g. Roberts, 1991, 1995). While areas that now receive higher protection from fishing [Flower Gardens (ID#5), Los Roques (ID# 16), Cayman (ID# 12), and Bonaire (ID# 17)] displayed slightly higher total biomass than areas with little protection (e.g., the Bahamas and western Caribbean), the difference was not significant. Since in some areas sites were surveyed within and outside of protected areas (e.g., Abaco (ID#1), Cayman (ID# 12), Virgin Islands (ID# 13), a more rigorous analysis would require assigning levels of protection at the site scale. However, assigning AGRRA sites with an unbiased estimate of protection from fisher-related mortality across the entire Caribbean is difficult because of the wide range of management and harvesting practices in place. An unbiased measure of total fish extraction, broken down by species and area (preferably at a fine spatial scale), is needed before fish density and biomass patterns related to fishing can be examined robustly at large spatial scales. Results of the 1-way ANOVA analysis using Bryant et al.’s (1998) “overexploitation threat index” show no clear pattern for total biomasss or herbivore biomass or carnivore biomass and can be interpreted to suggest several possibilities. For example, the modeling of overexploitation threat may not accurately portray fishing pressures on reefs. In fact, this is likely since the threat index is based on a global-level analysis using a reefs proximity to coastal settlements which vary greatly in size and in the proportion of people that harvest reef fishes. Not factored in the index as other factors that can influence fish populations are habitat quality, adjacent habitat availability (shelf area), and levels of management. It is also likely that AGRRA total fish density (and biomass) data are not very sensitive to the type of fishing pressure common at many of the areas assessed in this volume because they are heavily weighted by herbivorous fishes which, except for some large parrotfish such as Scarus guacamaia, are seldom heavily targeted. One of the most widely discussed examples in which nearly all reef fishes are harvested is Jamaica (e.g., Koslow et al., 1994). Indeed, during AGRRA assessments 39 conducted along the north coast of Jamaica in August 2000, total biomass (<1,500 g/100 m ), herbivore biomass (<1000 g/100 m ) and carnivore biomasss (<150 g/100 m ) were all significantly lower than had been found at any of the areas in this synthesis (Klomp et ah, in press). It would appear that the herbivore (and total fish) density (and biomass) indicators are only strongly affected by intensive overfishing, and that a regional signal associated with more selective targeted fishing may be more difficult to detect. Carnivore density (as well as size and biomasses) is considered a more sensitive indicator of the type of fishing pressure occurring at many of the areas in this synthesis. As discussed earlier, the high densities (and biomass) reported in the Abrolhos (ID#20) are thought to arise primarily from differences in both habitat type and assessment methodology. For the remaining areas, spatial patterns are apparent and suggest that portions of the western (ID#s7-9) and eastern (ID#sl3-15) Caribbean are consistently low in carnivore densities while portions of the southern Caribbean (ID#sl6-18) are consistently high. The overall low number of sightings for larger-bodied groupers and snappers (<~1/100 m ) as a whole suggests the entire region is overfished for many of these more heavily targeted species. Interestingly, the Bahamas subregion in particular Andros (ID#2), has a much higher proportion of large-bodied groupers {E. striatus, M. tigris, M. venenosa) to smaller-bodied species (E. cruentatus, E. guttatus, E. fulvus) than the rest of the region. The Bahamas subregion as a whole may have the least exploited grouper populations in the western Atlantic, an observation also supported by the large number of active grouper spawning aggregations (Sadovy, 1999). The lack of a significant relationship between carnivore density (or biomass) and overexploitation threat can be explained in part to the way this threat was modeled and in part also to the observation that the assessment areas probably do not represent a full gradient between “pristine” (unfished) and extremely overfished. In addition, the fish AGRRA belt transects are not considered the ideal size for quantifying the larger, solitary carnivores that are targeted by light-to-moderate fishing intensities. The narrow width and small areal coverage per site (600 m ) can often result in observers missing cryptic and/or shy serranids and lutjanids that are present on the reef Statistical power is low because of the overall low number of observations. Fish biomass reports based on Bohnsack- Bannerot cylinders are generally lower than those derived from belt transects, although, to my knowledge, no systematic comparison has been conducted. In addition, the large length-class intervals used by the observers reduce the power of the methodology to detect small differences in fish size. Ideally, belt transect data should be considered in conjunction with the Roving Diver sighting frequency data, as suggested by Schmitt et al. (2002). Although not examined in this synthesis, many of the individual papers in this volume include sighting frequencies derived from Roving Diver surveys. Based on species accumulation curves for three sites, Nemeth et al. (this volume) determined that a minimum of six Roving Diver surveys (or greater than seven hours search time) would be needed to approximate the actual species diversity in the deep St. Croix reefs. Flowever, much less time would be required to determine the presence and qualitative density of the 75 or so AGRRA species which could also be used to improve our ability to distinguish light-to-moderate levels of fishing pressure. 40 Algae and herbivory. The lack of any significant relationship between algal index and herbivore density/biomass in our data are somewhat in contrast to those of Williams and Polunin (2001) who reported that herbivore biomass showed a significant negative correlation with macroalgal cover and a positive correlation to cropped substrate in five localities across the wider Caribbean. Differences in the assessment techniques for quantifying algal cover are thought to be responsible for these different results. The AGRRA technique scores algal abundance and canopy heights in areas where there is at least 80% algal cover and does not record the absolute abundance of macroaglae for the entire hard substrata of a reef In fact, the inverse relationship between live stony coral cover and macroalgal canopy height that is evident in the AGRRA data supports the notion that macroalgal canopy height may be a function of the available hard substratum. Many algal species are unpalatable for fishes and canopy heights may only correlate with herbivorous fish density at relatively low abundances where algal choices are limited. However, these results also suggest that there are probably other factors (e.g., environmental conditions limiting algal growth) contributing to each of these signals at the regional scale. Densities of Diadema antillarum were low at nearly all surveys except for Costa Rica and thus probably do not exert a strong influence over algal assemblages at this time. Reef Health That three groupings resulted from the similarity analysis (Fig. 15) was somewhat surprising since the indices employed have ecological functions spaning a range from the individual-to-community-to-ecosystem levels (Table 6). In part, the high degree of similarity within the groups (>70%) arises because several indicators reinforced one another (e.g., inverse relationship between coral cover/density and macroalgal index as discussed earlier). Clearly, the choice of indicators had strong bearing on the outcome of this type of similarity analysis. Indicators that displayed the highest large-scale variation (e.g., recent mortality, carnivore density) had more influence on the groupings than those that had minimal large-scale variation (e.g., old partial mortality, colony diameter) (Table 5). None of the areas in the “better” category had experienced severe damage to its stony corals from the 1998 ENSO event although high percentages of mottled and pale colonies were seen in both Bonaire (ID# 17) and the windward Netherlands Antilles (ID# 14) at the time of their assessment (February, 1999 and December 1999, respectively). In contrast, four of the six assessments in the “worse” category were strongly influenced by this event as was Belize (ID# 10) (see Peckol et al., this volume). However, the other two survey areas in the worse category [Costa Rica (ID# 19) and Abaco (ID#1)] are subject to high chronic stresses, as reflected mainly in their low coral density, low coral cover, and high macroalgal index. Most of the areas in the “worse” category also had subnormal densities of carnivores while those in the “better” category had above-normal densities. Patterns observed in the overall state of the region and inferred causes (natural versus anthropogenic) for these patterns have begun to emerge from this preliminary analysis. It was expected that coral reefs remote from human population centers might be more intact and in better condition than those adjacent to population centers where there is a higher potential for exposure to pollution, overfishing, nutrient emichment. 41 sedimentation, recreational diving, and anchor damage. Results from this synthesis suggest such may not be the case. Often more remote reefs showed as much, if not more, evidence of reef degradation than reefs closer to human coastal development. For example: Andros (ID#2) had high recent partial mortality of corals, high macroalgal index, and few fish; Anegada and Guana in the British Virgin Islands (ID# 13) had high prevalences of coral disease; at Los Roques (ID# 16), A. palmata in the shallow barrier was nearly 100% standing dead; and in Mouchoir Bank, Turks and Caicos Island (ID#4), macroalgal abundance was high. Fringing reefs on some highly populated islands [i.e., Bonaire (ID# 17), windward Netherlands Antilles (ID# 14)] seem to have avoided sea surface temperature “hotspots” while remote reefs in the Bahamas [(i.e., Andros, San Salvador (ID#3)] did not. The observations that remote reefs are in poor shape does not imply that human actions are not influencing reef condition; this has been demonstrated conclusively by numerous studies (e.g., Ginsburg, 1994). Rather, it suggests that the driving forces influencing reef condition across the region are complex and probably involve multiple sources operating over several spatial and temporal scales. Regional (e.g., Diadema dieoff, some coral diseases) and subregional (ENSO-driven bleaching and associated diseases) stressors superimposed on localized states (e.g., reef development, degree of overfishing, disturbance history, abiotic conditions), and the degree to which reefs have ecologically adapted to these states over the past several decades, is a more plausible explanation for the current patterns. It also implies that human “proximity” now has a global reach, given atmospheric build-ups and transport of pollutants. The driving forces of western Atlantic reef degradation are operating at large spatial scales and management efforts should be directed towards these same regional scales. A critical issue that has practical consequences on the outcome of any reef health assessment relates to the benchmark used for the rankings. In this preliminary analysis, departures from ecologic norms were used as the principal criteria forjudging health. However, departures from normality are not necessarily unhealthy since natural reef ecosystems are expected to experience routinely disturbance events such as hurricanes (Rogers, 1993). A more meaningful measure of reef health may thus be resilience (Rolling, 1973) but there is significant debate with respect to the time scales necessary to gauge reef resilience (Done et ah, 1996). Evaluating reef condition based principally on ecologic criteria without regard to other criteria (e.g., socioeconomic, abiotic parameters) leads to its own bias. For example, a heavily managed reef can be ecologically subnormal but may be judged healthy if it fulfils a designated purpose. Thus, while the intent here was to examine patterns of normality and classify survey areas based on their broad groupings, applying judgmental health labels to these groupings is probably premature. One criticism to the AGRRA approach is that indicators and values are largely based on what we know about the functioning of coral reefs today rather than the type of pristine natural reefs that may have existed in the past. What represents a “normal” or even “healthy” reef today may well be “degraded” with respect to an earlier baseline (Jackson, 1997). Recognizing that ecological baselines have shifted (Jackson, 1997, 2001; Greenstein et ah, 1998), the value of making multiple observations across multiple spatial scales that can approximate the average state for the region today is still very high. The initial AGRRA norms synthesized in this volume can now be used as yardsticks to 42 evaluate reef condition like those for human health (e.g., blood pressure, infant mortality, prevalence of heart disease, life expectancy). In addition, these observations can be used with certain limitations within the hypothesis-testing framework (Underwood, 2000). The results presented in this synthesis, and in the papers of this volume, represent a preliminary analysis of a large amount of data. As additional AGRRA assessments are undertaken, the regional norms will shift to become more representative of the region, particularly for shallow reefs. In addition, more powerful analyses will be possible as our classification and stratification of reef types improves and as more species-specific indices are incorporated into the dataset. CONCLUSIONS • A mean live coral cover of 26% in the deep sites suggests significant losses have occurred over the past several decades but that substantial amounts of coral remain. • Significant bleaching-induced mortality associated with the 1998 ENSO event was most apparent in the western Caribbean and Bahamas subregions. • Linkages between infectious diseases, bleaching, and recent mortality were evident and are thought to be a result of increased pathogen activity and sensitivity of corals to disease and mortality during periods of elevated sea surface temperatures. • Coral and fish community species composition across the entire region were most strongly influenced by environmental and biogeographic factors. • AGRRA data provides further evidence on the regional decline of acroporids, although some moderate occurrences of A. palm at a were identified and the historic range does not appear to be reduced or lost in either species. • Large and perhaps unprecedented disease and bleaching impacts on the M annularis species complex were documented by many of the AGRRA surveys. Given the importance of Montastraea as the current and historic principal frame builder of reef slopes, this decline is cause for great concern. • The overall low number of sightings for larger-bodied groupers and snappers (<~1/100 m ) as a whole suggests the entire region is overfished for many of these more heavily targeted species. • Herbivore density (or biomass) and macroalgal index were not related at the scale of the entire region suggesting that there are probably other factors (availability of substrate, abiotic factors) contributing to each of these signals. • More remote reefs showed as much evidence of reef degradation as reefs more proximal to human coastal development. • Driving forces influencing present-day reef condition across the region are complex, and probably involve multiple sources operating over several spatial and temporal scales. 43 ACKNOWLEDGMENTS The data used in this synthesis were collected and generously provided by many colleagues who share in Robert Ginsburg’s vision of the AGRRA Program and the need for developing a regional understanding of Caribbean reef condition. To all I am extremely grateful. The National Center for Caribbean Coral Reef Research (NCORE), Ocean Research and Education Foundation (ORE) and the National Fish and Wildlife, Foundation (NFWF) coral reef grants program (#2001-0036-001) provided support during the preparation of this manuscript. I want to thank Robert Ginsburg, Patricia Richards Kramer, Judy Lang, Bob Steneck, Terry Done and Pedro Alcolado for advice and many useful discussions. Beth Fisher, James Byrne, Karlisa Callwood and Eduardo Martinez assisted with formatting and verifying large amounts of data over the last several years. Ken Marks is thanked for managing the AGRRA database and developing many of the queries that were needed to summarize the data. I am particularly grateful to Patricia Richards Kramer and Judy Lang who provided many scientific and editorial comments that substantially improved this manuscript. REFERENCES Adey, W.H. 1978. Coral reef morphogenesis: a multidimensional model. Science 202:831-837. Aronson, R., and W. Precht 1997. Stasis, biological disturbance, and community structure of a Holocene coral reef. Paleobiology 23:326-346. Aronson, R. B., and W. F. Precht. 2001 . Evolutionary paleoecology of Caribbean coral reefs. Pp. 171-233. In: W.D. Allmon and D.J. Bottjer, (eds.). Evolutionary Paleoecology: the Ecological Context of Macroevolutionary Change. Columbia University Press, New York, New York, USA. Ault, T.R., and C.R. Johnson 1998. Spatially and temporally predictable fish communities on coral reefs. Ecological Monographs 68:25-50. Bak, R.P.M., and B.E. Luckhurst 1 980. Constancy and change in coral reef habitats along depth gradients at Cura9ao. Oecologia 47:145-155. Bak, R.P.M., and E.H. Meesters 1998. Coral population structure: The hidden information of colony size-frequency distributions. Marine Ecology Progress Series 162:301-306. Bak, R.M.P., and G. Nieuwland 1995. Long-term change in coral communities along depth gradients over leeward reefs in the Netherlands Antilles. Bulletine of Marine Science 56:609-619. Bellwood, D. R, 1988. On the use of visual survey methods for estimating reef fish standing stocks. Fishbyte 6:14-15. 44 Bryant, D., L. Burke, J. McManus and M. Spalding 1998. Reefs at Risk. A Map-Based Indicator of Threats to the World’s Coral Reefs. World Resources Institute. 56 pp. Bythell J.C., E.H. Gladfelter, and M. Bythell 1 993. Chronic and catastrophic natural mortality of three common Caribbean reef corals. Coral Reefs 12:143-152. Clarke, K.R. and R.N. Gorley 2001. PRIMER v5: User manual/tutorial. PRIMER-E, Plymouth, UK, 91pp Clark, T. 1997. Tissue regeneration rate of coral transplants in a wave exposed environment. Cape D’Aguilar, Hong Kong. Proceedings of the Eighth International Coral Reef Symposium 2:2069-2074. Coates, B., A. R. Jones, and R. J. Williams 2002. Is ‘Ecosystem Health’ a useful concept for coastal managers? Proceedings of the Coast to Coast Meeting. 4-8 November, 2002, Australia. Pp. 55-58. Cochrane, W.G. 1977. Sampling Techniques (Third edition). John Wiley & Sons, New York, New York. 428 pp. Cortes, J., and M.J. Risk 1985. A reef under siltation stress: Cahuita, Costa Rica. Bulletin of Marine Science: 36:339-356. Costanza, R. 1992. Toward an operational definition of ecosystem health. Pp. 239-256. In: R. Costanza, B.G. Norton, and B.D. Haskell (eds.), Ecosystem Health: New Goals for Environmental Management. Island Press, Washington, D.C. Cowen, R.K., K.M.M. Ewiza, S. Sponaugle, C.B. Paris, and D.B. Olson 2000. Connectivity of marine populations: open or closed? Science 287:857-859. Done, T. 1992. Constancy and change in some Great Barrier Reef coral communities: 1980- 1990. American Zoologist 32: 655-662. Done, T.J. 1995. Ecological criteria for evaluating coral reefs and their implications for managers and researchers. Coral Reefs 14:183-192. Done, T.J., J.C. Ogden, W.J. Wiebe, and B.R. Rosen 1996. Biodiversity and ecosystem function of coral reefs. Pp. 393-429. In: H.A. Mooney, J.H. Cushman, E. Medina, O.E. Sala, and E.D. Schulze (eds.). Functional Roles of Biodiversity: a Global Perspective. John Wiley, New York, New York. Done, T. 1 997. Four performance indicators for intergrated reef resoursces management. Workshop on Intergrated Reef Resources Mangaement in the Maldives 1997:237-251. Dustan, P., and J.C. Halas 1987. Changes in the reef-coral community of Carysfort Reef, Key Eargo, Florida: 1974 to 1982. Coral Reefs 6:9\-W)6. 45 Diiyl, F.C. van 1982. Distribution of Acropora palmata and Acropora cervicornis along the leeward coasts of Cura9ao and Bonaire. 4th meeting Soc. Reef Studies, Leiden (Abstract). Edmunds P.J. 1991. Extent and effect of black band disease on Caribbean reefs. Coral Reefs 10:161-165. Edmunds, P.J., and R.C. Carpenter 2001. Recovery of Diadema antillarum reduces macroalgal cover and increases abundance of juvenile corals on a Caribbean reef. Proceedings of the National Academy of Sciences 98:5067-5071. Gardner, T. 2002. Coral Reefs of the Tropical Western Atlantic: a Quantitative Summary of Recent Temporal Change and the Relative Importance of Hurricane Impacts. M.S. University of East Anglia, Norwich. 94 pp. Geister, J. 1977. The influence of wave exposure on the ecological zonation of Caribbean coral reefs. Proceedings of the Third International Coral Reef Symposium., University of Miami, Florida 1 :23-30. Ginsburg, R.N., E. Gischler, and W.E. Kiene 2001 . Partial mortality of massive reef-building corals: an index of patch reef condition, Florida Reef Tract. Bulletin of Marine Science 69:1 149-1 173. Ginsburg, R.N., and P.W. Glynn 1994. Summary of the colloquium and forum. Pp. i-ix. In: R.N. Ginsburg (compiler). Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards and History, 1993. Rosenstiel School of Marine and Atmospheric Science. Miami. FL. Gladfelter, W.B. 1982. White band disease m Acropora palmata: implications for the structure and function of shallow reefs. Bulletin of Marine Science 32:639-643. Glynn, P.W. 1996. Coral reef bleaching: Facts, hypotheses and implications. Global Change Biology 2:495-509. Goreau, T.F. 1959. The ecology of Jamaican coral reefs. I. Species composition and zonation. Ecology 40:67-90. Green, E., and A.W. Bruckner 2000. The significance of coral disease epizootiology for coral reef conservation. Biological Conservation 96:347-361. Greenstein, B.J., H.A. Curran, and J.M. Pandolfi 1998. Shifting ecological baselines and the demise of Acropora cervicornis in the western North Atlantic and Caribbean Province: a Pleistocene perspective. Coral Reefs 17:249-261. 46 Harvell C.D., K. Kim, J.M. Burkholder, R.R. Colwell, P.R. Epstein, DJ. Grimes, E.E. Hofmann, E.K. Lipp, A.D.M.E. Osterhaus, R.M. Overstreet, J.W. Porter, G.W. Smith, and G.R. Vasta 1999. Emerging marine diseases - climate links and anthropogenic factors. Science 285:1505-1510. Hixon, M.A. 1991. Predation as a process structuring coral-reef fish communities. Pp 475-508. In P.F. Sale (ed.). The Ecology of Fishes on Coral Reefs. Academic Press, San Diego, California, USA. Holling, C.S. 1 973. Resilience and stability of ecological systems. Annual Reviews of Ecology and Systematics 4:1-23. Hubbard, D., and D.K Scaturo 1985. Growth rates of seven species of scleractinian corals from Cane Bay and Salt River, St. Croix, USVI. Bulletin of Marine Science 36:325-338. Hughes, T.P. 1984. Population dynamics based on individual size rather than age: a general model with a reef coral example. American Naturalist 123:778-795. Hughes, T. 1993. Disturbance: effects on coral reef dynamics. Coral Reefs 12:1 17-233. Hughes, T.P. 1994. Catastrophes, phase shifts, and large-scale degradation of a Caribbean coral reef. Science 265:1547-1551. Hughes, T.P., and J.H. Connell 1999. Multiple stressors on coral reefs: A long term perspective. Limnology and Oceanography 44:932-940. Hughes, T.P., and J.B.C. Jackson 1980. Do corals lie about their age? Some demographic consequences of partial mortality, fission, and fusion. Science 209:713-715. Hughes, T.P., and J.B.C. Jackson 1985. Population dynamics and life histories of foliaceous corals. Ecological Monographs 55:141-166. Jackson, J.B.C. 1997. Reefs since Columbus. Coral Reefs 16:S23-S32. Jackson, J.B.C. 2001 . What was natural in the coastal oceans? Proceedings of the National Acadamy of Sciences of the United States of America 98:541 1-5418. Jones, G.P., and C. Syms 1998. Disturbance, habitat structure and the ecology of fishes on coral reefs. Australian Journal of Ecology 23: 287-297. Kjerfve, B. and the CARICOMP Steering Committee 1998. CARICOMP: a Caribbean network of marine laboratories, parks, and reserves for coastal monitoring and scientific collaboration. Pp. 1-16. In: B. Kjerfve (ed.), CARICOMP-Carribean Coral Reef Seagrass, and Mangrove Sites. Coastal region and small island papers 3, UNESCO, Paris. 47 Klomp, K.D., K. Clark, K. Marks, and M. Miller In press. Condition of reef fish on Jamaica’s north coast signals late stages of overexploitation. Proceedings of the Gulf and Caribbean Fisheries Institute Fifty-fourth Annual Meeting. Koslow, J.A., K. Aiken, S. Anil, and A. Clementson 1 994. Catch and effort analysis of reef fisheries of Jamaica and Belize. Fishery Bulletin 92:737-747. Kramer, P.A., and P.R. Kramer 2002. Transient and lethal effects of the 1998 coral bleaching event on the Mesoamerican reef system. Proceedings of the Ninth International Coral Reef Symposium 2:1 175-1 180. Kuta, K.G., and L. Richardson 1996. Abundance and distribution of black band disease of corals in the northern Florida Keys. Coral Reefs 15:219-223. Lessios, H.A., D.R. Robertson, and J.D. Cubit 1984. Spread of Diadema mass mortality through the Caribbean. Science 35:335-337. Lewis, S.M. 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Lindquist, D.G., and M.R. Gilligan 1986. Distribution and relative abundance of butterflyfishes and angelfishes across a lagoon and barrier reef, Andros Island, Bahamas. Northeast Gulf Science 8:23- 30. Luckhurst, B.E., and K. Luckhurst 1978. Analysis of the influence of substrate variables on coral reef fish communities. Marine Biology 49:317-323. McClanahan, T.R., and N.A. Muthiga 1998. An ecological shift in a remote coral atoll of Belize over 25 years. Environmental Conservation 25:122-130. McClanahan, T.R., T. Kamukuru, A. Muthiga, M.G. Yebio, and D. Obura 1996. Effect of sea urchin reductions on algae, coral, and fish populations. Conservation Biology 10:136-154. Meesters, E.H., 1. Wesseling, and R.P.M. Bak 1 996. Partial mortality in three species of reef-building coral and the relation with colony morphology. Bulletin of Marine Science 58:838-852. Meesters, E.H., I. Wesseling, and R.P.M. Bak 1997. Coral colony tissue damage in six species of reef-building corals: partial mortality in relation with depth and surface area. Journal of Sea Research 37:131-144. Munday, P.L. 2002. Does habitat availability determine geographical-scale abundances of coral- dwelling fishes? Coral Reefs 21:105-116. 48 Munro, J.L., and D. Williams 1985. Assessment and management of coral reef fisheries; biological, environmental and socio-economic aspects: seminar topic. Proceedings of the Fifth International Coral Reef Congress 4:475-578. Porter, J.W. (editor) 200 1 . The Ecology and Etiology of Newly Emerging Marine Diseases. Hydrobiologia Vol. 460. Kluwer Academic Publishers, The Netherlands. Pp. 228. Porter, J.W., and O.W. Meier 1 992. Quantification of loss and change in Floridian reef coral populations. American Zoologist 23:625-640. Rapport, D.J., C. Gaudet, J.R. Karr, J.S. Baron, C. Bohlen, W. Jackson, B. Jones, R.J. Naiman, B. Norton, and M.M. Pollock 1998. Evaluating landscape health: intergrating societal goals and biophysical process. Journal of Environmental Management 53:1-15. Rapport, D.J., N. Christensen, J.R. Karr, and G.P. Patil 1999. The centrality of ecosystem health in achieving sustainability in the 2E‘ century: concepts and new approaches to environmental management. Transactions of the Royal Society of Canada, Series VI, Volume IX:3-40. Richardson, L. 1998. Coral diseases: what is really known? Trends in Ecology and Evolution 13:438-443. Roberts, C.M. 1991. Larval mortality and the composition of coral reef fish communities. Trends in Ecology and Evolution 6:83-87. Roberts, C.M. 1995. Effects of fishing on the ecosystem structure of coral reefs. Conservation Biology. 9:988-995. Roberts, C.M. 1997. Connectivity and management of Caribbean coral reefs. Science 278:1454-1457. Roberts, C.M., C.J. McClean, J.E.N. Veron, J.P. Hawkins, G.R. Allen, D.E. McAllister, C.G. Mittermeier, F.W. Schueler, M. Spalding, F. Wells, C. Vynne, and T.B. Werner 2002. Marine biodiversity hotspots and conservation priortites for tropical reefs. Science 295:1280-1284. Rogers, C.S. 1993. Hurricanes and coral reefs: the intermediate disturbance hypothesis revisited. Coral Reefs 12:127-137. Rogers, C.S., H.C. Fritz, III, M. Gilnack, J. Beets, and J. Hardin 1986. Scleractinian coral recruitment patterns at Salt River submarine canyon, St. Croix, U.S. Virgin Islands. Coral Reefs 3:69-76. Rosenberg, E., and Y. Ben-Haim 2002. Microbial diseases of corals and global warming. Environmental Microbiology 4:3 1 8-326. 49 Russ, G.R., and A.C. Alcala 1989. Effects of intense fishing pressure on an assemblage of coral reef fishes. Marine Ecology Progress Series 56:13-27. Sadovy, Y. 1999 Qggg Qf disappearing grouper: Epinephelus striatus, the Nassau grouper, in the Caribbean and western Atlantic. Proceedings of the Gulf and Caribbean Fishery Institute 45:5-22. Sale, P.F. 1998. Appropriate spatial scales for studies of reef fish ecology. Australian Journal of Ecology 23:202-208. Schmitt, E.F., R.D. Sluka, and K.M. Sullivan-Sealey 2002. Evaluating the use of roving diver and transect surveys to assess the coral reef assemblage off southeastern Hispanola. Coral Reefs 21 :2 16-223. Shinn, E.A., C.D. Reich, T.D. Flickey, and B.H. Lidz In press. Staghorn tempestites in the Florida Keys. Coral Reefs. StatSoft, Inc., 2002. STATISTICA Base Version 6.0. http://www.statsoft.com/ Steneck, R.S., and M. Dethier 1994. A functional-group approach to the structure of algal-dominated communities. Oikos 69:476-498. Syms, C., and G.P. Jones 2000. Disturbance, habitat structure, and the dynamics of a coral reef fish community. Ecology 81:2714-2729. Szmant, A.M., and N.J. Gasman 1990. The effects of prolonged “bleaching” on the tissue biomass and reproduction of the reef coral Montastrea annularis. Coral Reefs 8:217-224. Underwood, A.J., M.G. Chapman, and S.D. Connell 2000. Observations in ecology: you can’t make progress on processes without understanding the patterns. Journal of Experimental Marine Biology and Ecology 250:91 -\\5. Wheaton, J., W.C. Jaap, J.W. Porter, V. Kosminyn, K. Hackett, M. Lybolt, M.K. Callahan, J. Kidney, S. Kupfnel, C. Tsokos, and G. Yanev 2001 . EPA/FKNMS Coral Reef Monitoring Project Executive Summary 2001 FKNMS Symposium: An Ecosystem Report Card. Washington DC. Wilkinson, C. 2000. The 1997-1998 mass bleaching event around the world. Pp. 15-38. In: C. Wilkinson (ed.), Status of Coral Reefs of the World: 1998, Australian Institute of Marine Science, Cape Ferguson, Queensland. Williams, I.D., and N.V.C. Polunin 2001 . Large-scale associations between macroalgal cover and grazer biomass on mid-depth reefs in the Caribbean. Coral Reefs 19:358-366. Willis, T.J. 2001 . Visual census methods underestimate density and diversity of cryptic reef fishes. Journal of Fish Biology 59:1408-141 1. 50 Table 1. Summary site information for the 20 AGRRA assessments in this volume. ID# AGRRA Survey Modeled Sites (#) Reef typ e Other Region Area Date(s) Threat Benthic Fish Shallow (<5 m) Deep (>5 m) Index' •o c £ o u Over-exploitation Shallow (<5 m) Deep (>5 m) Total Shallow (<5 m) Deep (>5 m) Total Sampling unit (m)^ Reef crest Back reef Patch reef pther Reef slope Bank reef Other Diadema density (#/100 m^) 1 Bahamas Abaco, Bahamas Aug., 1999 H H 2 6 8 2 6 8 30x2 1 1 6 0 2 Bahamas Andros, Bahamas Aug., 1998 M M 15 16 31 15 16 31 30x2 15 16 0.6 ± 1.2 3 Bahamas San Salvador, June, 1998 M L 4 7 11 4 7 11 7.5 2 2 3 4 Bahamas 4 Bahamas Turks & Caicos Aug., 1999 M L 2 26 28 2 26 28 30x2 I 1 24 2 0 Islands 5 Gulf of Flower Garden Nov., 1999 L L 0 2 2 0 2 2 30x2 2 1.1 ±0.4 Mexico Banks, USA 6 Gulf of Veracruz, July, 1999 H H 3 3 6 — — — — 3 3 — Mexico Mexico 7 Western Yucatan, June - Sept,, M L 0 24 24 0 24 24 50x2 24 0 Caribbean Mexico 1999 8 Western Akumal/Xcalak, March, 1999 M L 2 12 14 0 6 6 30x2 2 12 0.2 ±0.9 Canbbean Mexico 9 Western Lighthouse July - Oct., M L 0 11 II 0 11 II 30x2 II 0 Caribbean Atoll, Belize 1999 10 Western Caribbean Belize May, 1999 M L 3 10 13 3 10 13 7.5 3 5 5 ... II Central Mdria La Gorda, July, 1999 M M 0 4 4 0 4 4 50x2 4 L8± 1.9 Caribbean Cuba 12 Central Cayman June, 1999, M M 2 40 42 2 40 42 30x2 1 1 39 1 0.2 ± 1.4 Caribbean Islands & June, 2000 13 Eastern Virgin July, 1999- H H 4 26 30 4 23 27 30x2 4 23 2 1 2.6 ±5 Canbbean Islands Aug., 2000 14 Eastern Windward Dec., 1999 M L 0 24 24 0 24 24 30x2 21 3 0 Caribbean Netherlands Antilles 15 Eastern Grenadines, June, 1999 M M 2 3 5 2 3 5 30x2 2 3 2.4 ±3 Caribbean St. Vincent 16 Southern Los Roques, Oct., 1999 L L 7 6 13 7 6 13 30x2 4 1 2 6 5.5 ± 10.8 Caribbean Venezuela 17 Southern Caribbean Bonaire Feb., 1999 M M 0 6 6 0 4 4 30x2 6 0 18 Southern Curasao Aug., 1998 H H 0 14 14 0 4 4 30x2 14 0 Canbbean & Jan., 2000 19 Southern Caribbean Costa Rica Oct., 1999 H M 1 2 3 1 2 3 30x2 1 1 1 23 ±33 20 Brazilian Abrolhos, Feb. -April, M L 6 7 13 6 5 11 30x2 6 7 0 Brazil 2000 Total 53 249 302 48 223 276 23 4 10 16 221 9 19 2.3 ± 5.6 'The threat index (H=high, M=medium, L=low) for each area is based on model outputs from Bryant et al.’s (1998) global Reefs at Risk classification. ^30 X 2 = 30 m long x 2 m wide belt transects; 50 x 2 = 50 m long x 2 m wide belt transects; 7.5 = 7.5 m radius cylinders 51 Table 2. Summary of mean sampling effort in deep sites for each assessment. ID# Mean depth (m) Mean sampling effort/site in deep (> 5m) sites (for a maximum of 1 0 transects) Benthic component (#) Fish component-belt transects Transects Corals Quadrats # Area (m^) 1 8 10.0 32 42 10.0 600 2 10 9.9 84 43 9.9 596 3 9 9.3 58 42 — — 4 14 9.6 91 48 9.8 591 5 21 10.0 118 48 10.0 600 6 10.5 10.0 50 41 — — 7 10.5 9.6 40 48 6.0 600 8 13 9.2 45 43 3.0 180 9 7.5 9.2 83 45 10.0 600 10 10 10.0 100 19 — — 11 7.5 10.0 84 43 6.0 600 12 10.5 9.6 90 48 9.7 582 13 10.5 9.8 66 47 10.0 600 14 13 9.5 55 40 10.0 600 15 10.5 10.0 96 49 10.0 600 16 10 5.8 56 18 9.8 590 17 10 8.0 95 32 9.8 585 18 13.5 8.4 138 19 3.3 195 19 6.5 10.0 47 48 10.0 600 20 6 9.6 88 48 10.0 600 52 Table 3. Summary of means and standard deviations for selected benthic AGRRA indicators for the 20 assessments (for a maximum of 10 transects/site). Stony coral density Partial-colony mortality (%) Stony corals (%) % Q Mean depth (m) >25 cm stony corals (#) Live stony coral cover (25%) >25 cm corals (#/10m) <2 cm corals (#/ mb c 0) o ai "O O Standing dead TD (L> JZ u C3 ■o u c/) C3 C/5 5 1 ^Igal overgrowth (A D 15 JZ c/5 tZ Snails/Worms Damselfish gardens A. Shall 1 ow (<5 m) sites 4 109 18.5 ±10.6 5.8 ±2.3 4.5± 1.3 2.0 ±8.3 24.1 ±30.6 11.0 0.9 2.8 2 1 1275 37.8 ± 16.7 10.0 ±3.2 2.5 ± 1.3 4.4 ± 11.7 21.6 ±25.4 8.3 14.3 5.9 4.2 5.1 8.5 2.2 3 4 165 ... 5.5 ±2.2 ... 0.8 ±3.8 28.8 ±28.7 15.2 1.8 1.2 0.6 ... ... ... 4 3 137 9.5 ±2.6 7.5 ±2.0 2.3 ±2.7 1.3 ±7.7 42.3 ±34.6 24.8 0 4.4 ... ... ... ... 5 ... ... ... ... ... ... ... ... ... ... ... ... ... 6 4.5 137 18.7±9.3 6.0 ±2.8 1.0 ±0.7 0.4 ±2.9 6.9± 11.8 ... 5.8 0 35.8 1.5 ... ... 7 ... ... ... ... ... ... ... ... ... ... ... ... ... ... 8 2 134 19.2± 11.8 7.5 ±2.4 2.7 ±3.3 26.9 ±22.9 37.8 ±23.3 6.0 10.4 13.4 0 6.7 27.6 0 9 ... ... ... ... ... ... ... ... ... ... ... ... ... — 10 4 292 ... 11.1 ±5.7 ... 3.4 ± 10.5 43.1 ±31.4 17 1 20.2 1 1.4 ... ... ... 11 ... ... ... ... ... ... ... ... ... ... ... ... ... ... 12 3.5 104 19.3±6.1 5.8 ±2.4 4.5 ±3.3 0.5 ±3.3 16.9 ±22.7 ... 0 4.8 0 0 3.8 1.0 13 4.5 248 14.0 ±7.2 8.0 ±3.3 4.2 ±2.7 3.6 ±5.3 44.8 ±33.9 1.6 28.6 2.8 37.1 17.3 25.8 2.0 14 ... ... ... ... ... ... ... ... ... ... ... ... ... ... 15 3.5 160 29.5 ±5.7 8.3 ± 1.4 3.1 ±2.9 1.9 ±7.2 22.1 ±21.5 6.3 2.5 0 0.6 0.6 18.1 1.3 16 2.5 332 14.6 ± 12.4 8.1 ±4.0 1.3 ±2.0 5.4 ±9.8 21.1 ±25.3 52.1 3.6 0.9 2.1 1.2 21.4 0.6 17 ... ... ... ... ... ... ... ... ... ... ... ... ... ... 18 — — — ... ... ... ... ... ... ... ... ... ... ... 19 2 44 3.4± 1.0 6.1 ±2.7 0.3 1.3 ±6.3 28.0 ±26.1 6.8 4.5 6.8 40.9 11.4 0 4.5 20 4 503 10.6 ±5.4 12.4 ±6.0 9.4 ±2.4 0.1 ±0.9 18.3±21.9 0 0.4 0 ... ... ... ... Mean 3.2 280 17.7 ±9.5 7.9 ±2.2 3.3 ±2.5 4± 7.1 27.4 ±11.6 14.9 7.2 3.4 12.3 5.5 15.0 1.7 B. Deep 1 (>5 m) sites 8 194 13. 5± 10.1 l.l± 1.7 4.7 ± 1.2 4.4 ± 12.4 26.1 ±28.7 6.2 14.9 1.0 8.2 1.0 2 10 1347 25.0 ± 14.3 10.1 ±4.8 9.1 ±4.2 16.5 ±29.9 17.9 ±24.3 4.1 8.5 17.7 7.9 0.1 0.1 2.9 3 9 406 ... 6.9 ±2.7 ... 2.9 ±8.8 29.5 ±29.7 11.6 3.2 0.5 1.5 1.7 0.2 1.2 4 14 2362 19.1 ± 8.7 10.2 ±3.7 5.9 ±2.7 3.3 ±11.8 24.2 ±25.2 1.1 0.3 5.4 ... ... ... ... 5 21 235 57. 8± 13.1 12.4 ±2.8 1.9±0.8 1.8±6.1 9.5 ± 16.4 1.3 3.4 0 3.4 7.2 ... 1.3 6 10.5 151 22.2 ± 13.0 6.9 ±3.4 3.1 ±0.1 0.1 ± 1.1 8.2 ± 13.6 ... 2.6 0 35.1 ... ... ... 7 10.5 951 14.5 ±6.9 5.3 ±2.3 3.5 ±2.7 17.8 ± 19.9 14.5 ± 16.4 5.6 10.4 13.9 ... ... ... 1.7 8 13 541 20.9± 13.1 6.1 ±2.4 1.7±1.1 5.0 ± 12.0 26.3 ±25.9 4.1 11.1 3.7 0.4 3.9 9.4 9 7.5 909 21.1 ±9.9 9.6±2.5 61 ±3.8 2.9 ±9.0 24.8 ±24.7 0.4 10.0 3.5 31.7 10.5 8.6 40.9 10 10 1004 ... 11 ±5 ... 6.3 ± 18.0 29.9 ±29.9 11.8 17.6 1.9 1.1 ... ... ... 11 7.5 334 20.2 ±5.1 9.1 ±2.6 1.7± 1.1 5.5 ± 12.8 29.4 ±28.1 1.8 5.4 4.8 1.2 ... ... 47.6 12 10.5 3617 20.5 ±9.1 10.2 ±4.2 3.9±2.1 4.2 ± 13.2 24.0 ±25.1 1.1 1.4 5.3 0.3 0.4 0.2 2.5 13 10.5 1723 22.3 ± 12.1 8.9 ±4.4 9.1 ±4.6 1.5 ±5.6 29.8 ±29.6 1.0 26.8 6.2 8.6 8.0 1.2 5.2 14 13 1309 21.4± 12.7 6.5 ±3.2 4.9 ±4.6 1.4 ±7.7 20.8 ±24.8 0.8 26.5 1.4 0.2 0.4 0.4 1.3 15 10.5 288 39.1 ±9.8 10.4 ±2.9 2.8 ±2.4 1.6 ±5.7 26.2 ±25.6 0.3 11.8 4.2 3.8 1.4 1.0 17.4 16 10 335 41.6±2.2 10.4 ±3.2 2.5 ±3.5 4.1 ± 10.1 20.0 ± 18.8 6.6 4.2 3.0 1.2 3.6 2.1 14.3 17 10 570 46.5 ± 13.3 13.3 ±4.1 4.2 ±2.7 0.7 ±2.5 31.3 ±27.5 2.3 27.2 5.3 1.1 6.5 0.4 13.3 18 13.5 1925 35.2 ± 11.9 17.9±5.3 4.0±4.1 1.1 ±3.6 20.0 ±20.4 0.5 0.1 10.4 0.8 1.5 0.7 3.5 19 6.5 93 2.5± 1.1 5.4± 1.8 3.5±3.1 1.8±7.1 14.3 ± 18.5 ... 9.7 9.7 22.6 2.2 ... ... 20 6 619 20.5 ±9.9 11.5±4.6 14.6 ±3.8 0.4 ±3.9 11.3 ± 18.6 0 2.9 0 ... ... ... 0.3 Mean 10.6 945.7 25.8 ± 13.3 9.3 ± 3.3 4.4 ±3.4 4.2 ±4.7 21.9 ±7.2 3.5 9.9 4.9 7.6 3.6 1.5 1.5 — = no data Table 4. Comparison of recent and old partial-colony mortality, incidence of standing dead, bleaching, and disease for the 20 most common stony coral species or taxa (all colonies >25 cm in diameter), for a maximum of 10 transects/site. 53 'BBD = black-band disease; WBD = white-band disease; WP = white plague; YBD = yellow-bloteh (=yellow-band) disease; Other = dark spots, white pox (=patchy necrosis), etc. Table 5. Comparison of the coefficient of variation (standard deviation divided by the mean) for 12 AGRRA indicators in the deep sites (>5 m) at four spatial scales. 54 55 S-i _o tJ) c o X (U T3 C ca »-i. > '£ o ^ a 3 -a o c < V " — C3 •2 E a t OB a c3 C _! C/D — , J W o ai $3 a ^ « o o C U a 4/ 2 « I ? tf5 W < ^ 0 ^ W) 9 5 u ^ ,3 OB ^ £ o u J O ■E 3 C ■O 3 '> E ^ E - tj 1) >- > -3 2 o o a > c2 iS c S OB, X ■E E w /"“v ^ s .2 ^ V ^ 3 O c s: X II ip ^ 1 o o u = 2 ti o o D c ck: = 5 cd /-% o- e "a 2 2 S O c -o C O i> CO > JJ X X 2 2 o o > c2 c o -5 £ u •E E w « CO _ > c2 S _ 9 OB, £ •E E w 2 Sn C/D ^ IX .H JJ X 2 ^ 2 o o > > c2 ,3 ^ S _ o oi), ^ £ c/5 sO 5 - u ^ X j x CQ § 1 •o O c/5 Urn O T3 r9 J3 ^ o ^ U« c/5 cd ^ o s ^ Oh {rt Cd 2 O X « -5 « X2 ^ 3 tu ^ c/5 > O 3 < -2 < ^ H X W CJ 00 C 3 II ? .2f o n- U s c « 3 c/5 s > 5 c/5 £ o o ^ u tu X 4> ^ X cd Cd > g ‘S 42 3 II [1 X ? .2f> O "T u u c u § 2 c o 3 40 ^ u '44 O nO d) ^ ■3 E '5^ cd o 1) ^ X oi o > P S .2^ O 30 J to u -J density (#/l 00 m^) High f.fayojabk A^’PJt'9?- AGRRA herbivore Community S I Chronic Low = unfavorable 31 1 5 - -54 Andros density (#/ 1 00 m^) Ecosystem Hi^h_ 7.fay_ojaWe_ Abacq AGRRA carnivore Community S. L S Either Low = unfavorable 6 0.4 - 26 St. Vincent density (#/100 m^) Ecosystem High = favorable Abrolhos 56 Plate 2A. Historically, many of the shallow crests of ocean-facing Caribbean reefs looked much like this luxuriant thicket of Acropora palmata. Beginning in the early 1970’s, colonies of palmata and of A. cervicornis began to die from the effects of what has become known as white-band disease. (Photo Robert N. Ginsburg) Plate 2B. The decline of acroporids has been so pervasive in the last three decades that numerous reef crests are now little more than vast cemeteries of dead coral. Many large dead acroporids are still “standing” in growth position (as shown here). AGRRA surveys have revealed a number of healthy reefs off Andros (this volume) and Cuba (subsequent assessments) where live acroporids are locally abundant. (Photo Robert S. Steneck) 57 Plate 3A. How much coral is in a coral reef limestone? The wall above in an abandoned quarry, now a State Park on Windley Key, Florida, shows a section of a Late Pleistocene reef limestone, ca. 125,000 years old. Measurements of the total area of coral that is visible to the naked eye within a five block area gave an average of 39% . Only the larger massive coral skeletons are in growth position, and most coral is fragmented. Including the smaller fragments and sand-sized grains of coral would increase the total to nearly 50% for this patch reef community. (Photo Robert N. Ginsburg) Plate 3B. AGRRA indicators can be used to provide norms of reef condition, like the value of calibrating the geological record of a coral community exemplified 3A. As additional AGRRA assessments are undertaken, the regional nonns will become more representative of the region thus moving towards the development of a regional baseline of current coral reef health. The combination of both past and current temporal perspectives is key in understanding how to preserve coral reefs for future generations. (Photo Bernhard Riegl) 58 f"'"^wFCS ^•^FCP Fowl Cay Preserve Man-O-War Cay • MOWSN , • MOWSS \ 'SR •ECN ECM • ECSO ECSI Tiloo Cay Pelican Cays Land and Sea Park SCB, SCF Sandy Cay LCN Lynyard Cay LCS km -• 26” 30’N IT 00’ W Figure 1. AGRRA survey sites offshore Abaco, Bahamas, with outlines for the Fowl Cay Preserve and Pelican Cays Land and Sea Park. Abbreviations: LCS = Lynyard Cay south, LCN = Lynyard Cay north, SCB = Sandy Cay backreef, SCF = Sandy Cay forereef, ECSI = Elbow Cay south inner, ECSO = Elbow Cay south outer, ECM = Elbow Cay middle, ECN = Elbow Cay north, SR = Storr’s Reef, MOWSS = Man O’ War Cay south of south channel, MOWSN = Man O’ War Cay north of south channel, FCP = Fowl Cay pinnacles, and FCS = Fowl Cay shallow. A RAPID ASSESSMENT OF CORAL REEFS NEAR HOPETOWN, ABACO ISLANDS, BAHAMAS (STONY CORALS AND ALGAE) BY JOSHUA S. FEINGOLD,’ SUSAN L. THORNTON,* KENNETH W. BANKS, ^ NANCY J. GASMAN,^ DAVID GILLIAM,* PAMELA FLETCHER,^ and CHRISTIAN AVILA* ABSTRACT Coral reefs at 1 3 sites ranging in depth from 1 - 1 6 m near Hopetown, Abaco Islands, Bahamas were surveyed utilizing the Atlantic and Gulf Rapid Reef Assessment (AGRRA) benthos protocol. A total of 35 species of scleractinian corals and 2 species of calcareous hydrocorals were observed. The overall coral cover averaged just over 14%. Among corals that were at least 10 cm in diameter, small colonies (<40 cm diameter) predominated in all sites except for the Fowl Cay pinnacles where 68% were larger than 60 cm in diameter. Large colonies (>40 cm diameter) were also found in the Lynyard Cay spur-and-groove formations and the Sandy Cay fore reef. Zero-4% of the colonies were affected by disease. Total (recent + old) partial-colony mortality ranged from 9-31% (both extreme values being found in outer reef crests). Turf algae were the most common algal functional group overall. Macroalgae were ubiquitous, however, with relative abundance values of about 25-47%. Macroalgal indices (a proxy for biomass) ranged from 64 in the Sandy Cay back reef to 184 in the Fowl Cay outer reef crest. INTRODUCTION The AGRRA protocol is being applied throughout the Bahamas and Caribbean to document the condition of reefs with a technique that allows interregional comparisons. This study targeted the reefs of the central Abaco Islands because of their location near the northeastemmost extension of the Bahamas platform. Reefs at the extremes of their geographic range occur near important physiological thresholds for stony corals and may respond earlier to global climate change or anthropogenic impacts than more centrally located reefs. With the exception of Bermuda, Abaco’s reefs are nearest the northern limit of shallow reef formation in the wider Caribbean. Additionally, the reefs of central Abaco are near an area of relatively high population density and user pressure from local and tourist fishers and divers. *National Coral Reef Institute, Nova Southeastern University Oceanographic Center, 8000 North Ocean Drive, Dania Beach, Florida 33004. Email: joshua@nova.edu 2 Broward County Department of Planning and Environmental Protection, 218 S.W. 1st Avenue, Fort Lauderdale, FL 33301. 60 The Abaco Islands archipelago, which is composed of middle-to-late Pleistocene eolianites, skirts the eastern margin of Little Bahama Bank (Locker, 1980). Great Abaco itself is of middle Pleistocene age and acts as a barrier to the eastward transport of fine sediments from the Bank interior. A shallow lagoon, the Sea of Abaco, is bounded by Great Abaco to the west and southwest. A string of windward cays to its east and northeast act as a barrier to waves. However, deep tidal channels can act as “energy windows” allowing energy flux (from open-ocean waves and tidal currents) into the lagoon. For example, at Sandy Cay which lies within the lagoon, a well-developed fringing reef thrives on a topographic high where it is exposed to open-ocean waves provided by the wide passes between the string of Lynyard Cay, the Pelican Cays (three islands), and Tiloo Cay. The reef community on the seaward side of Lynyard Cay also lies atop a topographic high; however, the steep slope of the narrow shelf may prevent the reef structure from being well developed (Kenneth Banks, personal observations). A relatively continuous bank-barrier reef is known to extend just offshore of the eastern cays of Abaco from the northern tip of the Little Bahama Bank, west of Walker’s Cay, southward to the middle of Elbow Cay near Hopetown, a distance of approximately 160 km. Southward of this point the shelf becomes narrow with a steep slope limiting the extent of suitable reef development area. However, spur-and-groove formations have developed at southern reefs offshore of Lynyard Cay on the topographic high spot. Storr (1964) previously described several lines of unusual parallel narrow linear reefs offshore of Johnny’s Cay, several kilometers north of Hopetown on Elbow Cay. Otherwise the Abacos’ reef flora and fauna have not been studied. Natural impacts on the reefs of Abaco are poorly documented. Frequent hurricane activity (e.g.. Hurricanes Erin, 1995, and Floyd, 1999) undoubtedly affects their community dynamics. Moreover, the extent of anthropogenic impacts is not currently known although Woodley et al. (2000) mentioned that overfishing, discharge of human wastewater, and loss of coastline habitat are contributing to degradation of marine ecosystems in the Bahamas. Two of the study areas, Sandy Cay and Fowl Cay, are within marine protected areas (Pelican Cays Land and Sea Park and Fowl Cay Bahamas National Trust Preserve, respectively) where harvesting of marine organisms is not permitted (Dodge, 1999; Woodley et al., 2000). Sandy Cay (in the park) and Lynyard Cay (south of the park and distant from population centers) appear to be relatively undisturbed by direct anthropogenic influences yet development is increasing in the Abacos archipelago. The parks currently do not have staff, and some poaching undoubtedly occurs, but many of the local dive charter operators attempt to aid in enforcement of the park rules. Additionally, all interviewed local inhabitants of Hopetown, including charter boat captains, dive operators, restaurateurs and resort operators, concur that these are “no take” areas. Charter fishing activities are generally targeted on pelagic fish species, or take place in the backwaters and tidal creeks of the Sea of Abaco, and probably don’t impact the reefs significantly. However, preliminary observations in November 1997 at locations of Storr’s (1964) study suggested that fish populations have been reduced (Kramer et al., personal communication). The northerly reefs (Man O’ War, Fowl, and Elbow Cay) surveyed in this investigation are in close proximity to the more developed areas around Marsh Harbour, Hopetown and Man O’ War Cay, locally known as the Hub of Abaco. The residents of this area depend heavily on commercial-scale harvesting of finfish (hook and line and spearguns), spiny lobster 61 (spearguns) and conch (collected by divers). Catches are exported to the United States or sold locally to supply the tourist industry. Hence, this project provides important data for more effective management of Great Abaco’s southern reefs and a baseline for future comparisons. METHODS From August 9-16, 1999 a team of seven divers trained in the AGRRA protocols surveyed reefs near Hopetown on Elbow Cay (Fig. 1). Aerial photographs and discussions with local residents, followed by ground-truthing, allowed the selection of representative reefs along a north-south geographical gradient within small-boat operating distance of Hopetown. The four sites at Sandy and Fowl Cays were strategically located in marine preserves. The Storr’s Reef site was located offshore of the linear reef forms studied by Storr (1964). At each site we focused on areas with the greatest reef development and highest cover of live scleractinian corals. Locations without good reef development (such as the southern end of Elbow Cay and Tiloo Cay) were not visited. Surveyed habitats included a well-developed fringing reef at Sandy Cay (the enormous thicket of dead Acropora palmata on its reef crest was not surveyed). The Sandy Cay back reef lacks extensive development of Acropora cervicornis but several patches were present in the surveys. Its fore reef, with massive corals on a steep slope extending from near sea level to a sandy plain at 10m depth, resembles comparable habitats in the bank-barrier reef. Surveys on the bank-barrier reef were conducted at: low-relief outer reef crests (e.g., Storr’s Reef and Elbow Cay north); low-relief spur-and-groove formations (Lynyard Cay); a shallow pinnacle reef (Elbow Cay south, outer site); and structurally complex pinnacles that rise from depths of 12-15 m to near the surface (Fowl Cay, where transects circumnavigated the deep pinnacle bases). The AGRRA benthos protocol Version 2 (see Appendix One, this volume) was followed with the following modifications. Stony corals with maximum widths of 10 cm or larger were included because Abaco’s live corals are typically of smaller sizes than those in other Caribbean locations and coral cover was low (personal observations). Coral diameters were measured to the nearest 5 cm. Sediments in the algal quadrats were fanned with quick hand motions prior to estimating the abundance of crustose coralline algae. Coral identifications were aided by reference to Humann’s (1996) field guide. Latitude and longitude were determined using a global positioning system (GPS) receiver, Magellan GPS 315. Readings were confirmed or updated as necessary in August 2000 after selective availability (SA) was turned off Quality assurance surveys of reference transects were initially performed at two sites (Hopetown Reef and Sandy Cay) by all team members who compared results to determine any differences regarding taxonomy, measurements, mortality, etc., and reach a consensus. 62 RESULTS Thirteen sites ranging in depth from 1-16 m were surveyed along the offshore shelf between Lynyard Cay in the south and Fowl Cay in the north (Fig. 1; Table 1). We found the reef morphology at most offshore sites to be typical of bank-barrier reefs (Goreau, 1959; Glynn, 1973). For example, the low-relief spur-and-groove system off Lynyard Cay runs normal to the prevailing wave assault at depths of 6-1 1 m. Most of these reefs were composed primarily of massive framework builders such as Montastraea spp. and Diploria spp. The branching elkhom coral, Acropora palmata, was not evident as a primary framework builder except in the Sandy Cay fore reef. However, the deep pinnacle reefs at Fowl Cay have numerous Agaricia agaricites and large (up to 400 cm diameter) colonies of Montastraea faveolata. Massive corals and Acropora cervicornis are found in the back reef of Sandy Cay. Stony Corals Density and percent cover. A total of 1,443 stony corals that were >10 cm in diameter were counted in 186 transects. Coral density averaged 7 colonies/ 10 m transect (0.7 colonies/m) when all sites were combined (Table 1). Density values ranged from 5.0- 8.5 colonies/10 m transect (-0.50-0. 85 colonies/m). Two of the three sites with the highest density were in the Sandy Cay protected area. Live stony coral cover, which averaged over all sites was just over 14% (mean=14.3, sd =7.6, n=13 sites), varied from about 8.5% in the Elbow Cay north outer reef crest to about 22.5% in the Sandy Cay fore reef (Table 1). Two of the three sites with the highest coral cover (Sandy Cay fore reef and Fowl Cay pinnacles) were in marine protected areas. The pooled average coral density in the four protected sites (Sandy and Fowl Cays) was slightly higher than the pooled density for the nine sites that are not protected (0.78 colonies/m versus 0.68 colonies/m). Similarly, the percent live coral cover averaged for the four protected sites (mean=16.4, sd=4.4) was somewhat greater than that for the nine that are not protected (mean=13.3, sd=2.8). Neither of these differences between protected and unprotected sites was significant (Rank Sum Test, Ambrose and Ambrose, 1995). Species. A total of 36 taxa of scleractinian corals and two species of calcareous hydrocorals were observed including species seen off the surveyed transects (Table 2). Porites astreoides was the most abundant species of the >10 cm corals in all sites (Fig. 2 A) with the exception of the deep Fowl Cay pinnacles where mounding Montastraea faveolata and foliose Agaricia agaricites predominated. Diploria clivosa and D. strigosa were also numerically abundant in the shallow (<6 m) outer reef-crest habitats (Fig. 2B) whereas the spur-and-groove formations at Lynyard Cay had high abundances of Montastraea faveolata and Siderastrea siderea (Fig. 2C). Millepora spp., Acropora palmata and Montastraea faveolata were regularly observed in most sites. Size. In most (11/13) sites, corals that were smaller than 40cm in diameter (largely comprised of Porites astreoides and Diploria spp.) made up the largest portion of the >10 cm coral community (Fig. 3A,B). Large corals were infrequently encountered with size classes of over 200 cm comprising 0-10% of the pooled surveyed colonies. At the Fowl 63 A All sites n = 1,322 corals B Low-relief reef crests n = 742 corals CN, DL, AT, DICH, AP = Acropora palmata AG = Agaricia agar idles, AT = A. tennifolia CN = Colpophyllia natans DEN = Dendrogyra cylindrus DICH = Dichocoenia stokesi DC = Dip lor ia clivosa, DL = D. labyrinthiformis, DS = D. strigosa MA = Montastraea annularis, MAE = M. faveolata, MC = M. cavernosa, MFR = M franksi MIL = Millepora alcicornis and M.complanata MM = Meandrina meandrites MYCET = Mycetophyllia spp. PA = Porites astreoides, PP = P. porites SI = Stephanocoenia intersepta SS = Siderastrea siderea C Spurs and grooves n = 177 corals AT, SI, MYCET, Figure 2. Species composition and mean relative abundance of all stony corals (>10 cm diameter) in (A) All Abaco, Bahamas sites, (B) Low-relief outer reef crests (7 sites), and (C) Spurs and grooves (2 sites). Coral species with < 1% are placed in a grouped category; Other in (A) = Agaricia tenuifolia, Colpophyllia natans, Dendrogyra cylindrus, Dichocoenia stokesii, Diploria labyrinthiformis, Madracis mirabilis, Manicina areolata, Meandrina meandrites, Montastraea franksi, Mycetophyllia spp., Porites porites, and Stephanocoenia intersepta. Cay pinnacles, however, smaller (<40 cm) corals were less than 20% of the total and nearly 20% of all colonies were larger than 150 cm in diameter. Mean colony diameters (Table 3) overall ranged between 32 cm (Man O’War Cay north of the south channel) and 62 cm (Sandy Cay fore reef). The largest size classes (> 100cm) were mainly comprised of Montastraea faveolata (Fowl Cay pinnacles) and Acropora palmata; small corals (10-20cm) were mostly represented by Porites astreoides, Diploria strigosa, Diploria clivosa and Agaricia agaricites. Recruits. Porites astreoides was the most eommon coral recruit. Together with Siderastrea radians and S. siderea, these three speeies accounted for nearly half (47%) of all recruits observed in the surveys (Fig. 4A). The “unidentified” category accounted for 21% and Manicina, Dichocoenia, Favia and Agaricia agaricites composed an additional 64 23%. The remainder (9%) were mostly reef builders that were only observed once or twice as recruits, including M. annularis, D. strigosa, D. clivosa, A. cervicornis, A. palmata, and two non-reef-building genera, Scolymia and Mycetophyllia. Similar recruit frequencies were observed in the low relief-reef crest sites (Fig. 4B), but Manicina recruits were more abundant than average at the spur and grove sites (Fig. 4C). Mean coral recruit density ranged from about 0.1-0.5/0.0625 m (~1.6-8 colonies/m ) (Table 4). A Low-relief outer reef crests 50 40 30 c (O I 20 [ I Porites astnoides n = 1 58 corals Diploria clivosa n = 1 50 corals H D. strigosa n = 1 26 corals 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-99 Diameter (cm) Figure 3. Size-frequency distributions of >10 cm diameter colonies of (A) Porites astreoides, Diploria clivosa and D. strigosa in low-relief outer reef crests (7 sites) and (B) Porites astreoides in spurs and grooves (2 sites) off Abaco, Bahamas. A All sites n = 181 corals AC = Acropora cervicornis AG = Agaricia agaricites DICH = Dichocoenia DC = Diploria clivosa DS = Diploria strigosa FF = Favia fragum MAN = Manicina areolata MYCET = Mycetophyllia PA = Porites astreoides SR = Siderastrea radians SS = Siderastrea siderea UNID = Diploria spp., Mycetophyllia spp., Scolymia spp., and/or Montastraea annularis species complex 65 Figure 4. Species composition and mean relative abundance of all stony coral recruits (>2 cm diameter) in (A) All Abaco, Bahamas sites, (B) Low-relief outer reef crests (7 sites), and (C) Spurs and grooves (2 sites). Coral species with <1% are placed in a grouped category; other in (A) = Acropora palmata, Diploria clivosa, Montastraea annularis, Mycetophyllia, Scolymia. Stony Coral Condition Partial Mortality. Recent partial-colony mortality (hereafter recent mortality) of the >10 cm stony corals was lowest (nearly zero) in the Sandy Cay back reef and averaged <3% in eight sites (Table 3). The three lowest mean values for old partial- colony mortality (hereafter old mortality) and total (recent and old) mortality (hereafter total mortality) occurred in the two outer reef crests near Man O’ War Cay and in the Sandy Cay back reef (~8-16% and 9-16%, respectively). The highest values for both recent mortality (6%) and total mortality (31%) were found in the Elbow Cay north outer 66 reef crest. Similar patterns of mortality were displayed on corals in the low-relief reef crest and those in the spur-and-groove reefs (Fig. 5A,B)- However, more colonies were affected by old mortality than by recent mortality. Also, a greater percentage of the recent mortality occurred on a smaller portion of the affected corals. Old mortality showed a more uniform distribution with respect to amount of the colony affected. Overall, an average of 3.6% of all the large stony corals were “standing dead” (100% total mortality, still in growth position). Three sites (both near Man O’ War Cay and Fowl Cay pinnacles) had no standing dead colonies while the highest value (8%) was observed in the Elbow Cay north site (Table 3). The species most frequently scored as standing dead were Acropora palmata, Diploria strigosa, Millepora complanata, M. alcicornis and Montastraea annularis. Diseases. The overall incidence of coral disease was low, with only 0-1% of all colonies being affected in 1 1/13 sites (Table 3). Diseased colonies were somewhat more abundant in the Fowl Cay pinnacles (2.5%) and the Sandy Cay fore reef (4%). The most commonly affected species included Diploria strigosa, Montastraea annularis and Montastraea faveolata. Black-band disease affected one surveyed colony each of M. faveolata, Diploria clivosa and Colpophyllia natans and it was noted on two colonies of Siderastrea siderea in areas not covered by the transects. Yellow-blotch disease (three colonies) and white plague (two colonies) were also observed. Several other surveyed colonies exhibited tissue sloughing or bands or borders of recently exposed skeleton surrounding live tissues. Due to the unclear etiology it was not possible for us to identify the putative disease agent in these cases and they were classified as “unknown.” Because the transects did not intercept many of the colonies of Diploria strigosa that were sloughing tissues, a haphazard swimming survey was performed in two sites (Elbow Cay middle, outer reef crest and Lynyard Cay north, spurs and grooves). All colonies encountered that were greater than 20 cm in diameter (with some as large as 55 cm) were examined. Of 79 surveyed corals, 26 were dead, 34 were observed to be actively sloughing tissue or with recent partial mortality, and only 19 appeared “healthy.” Additional perturbations. Other incidents of mortality observed in the transects included overgrowth of several coral species by the boring sponge Cliona (two colonies of Porites astreoides and one colony each of Diploria labyrinthiformis , Montastraea cavernosa, Porites porites, and Siderastrea siderea). The hydrocoral, Millepora alcicornis was observed to overgrow two colonies of Porites astreoides and parrotfish bite-marks were seen on a colony of Acropora palmata. Relatively few (0-3%) corals were bleached (Table 3), with entire colonies exhibiting very pale to white tissues. Abundance of Algae and Diadema antillarum Seven hundred fifty-four algal quadrats were examined during the survey. Macroalgal relative abundance values ranged from less than 30% in six sites to nearly 50% in the Elbow Cay north outer reef crest (Table 4). Mean macroalgal heights varied between about 2 cm in the Sandy Cay back reef and 5.5 cm in the Fowl Cay outer reef crest where macroalgal indices (macroalgal relative abundance x macroalgal height, a proxy for macroalgal biomass) were also lowest (64) and highest (184), respectively. 67 c 10 cm diameter) in (A) Low-relief outer reef-crests (7 sites) and (B) Spurs and grooves (2 sites) off Abaco, Bahamas. 68 By a large margin, the highest relative abundance of crustose coralline algae (45%) was in the Sandy Cay fore reef (the only site in which they predominated), whereas the lowest was in the outer reef crest at the Man O’ War south of south channel site (10%). Overall, turf algae constituted the most abundant algal functional group in the two spur-and- groove sites off Lynyard Cay and in many (5/7) of the low-relief outer reef crests. Only a few (<5) individuals of the long-spine sea urchin, Diadema antillarum, were seen during underwater operations in Abaco and none was observed in the invertebrate belt transect surveys (Table 4). DISCUSSION Coral cover in Abaco was generally similar to that observed further south in the Exuma Cays by Chiappone et al. (1997). The difference in latitude between Abaco and the Exumas is 3° or approximately 300 km. Chiappone et al. (1997) reported average live stony coral coverage of 22.8% for fringing reefs and 1 1.2% for windward hard-bottom habitats. In our surveys we found mean coverage of 18.5% in the Sandy Cay fringing reef and 13.5% for all other Abaco sites combined. The relative abundance of species in the studies are different, however, and one should note that the Exuma study included patch reefs (hosting Montastraea annularis, not widely documented in this survey) and channel reefs, neither of which type of habitat was examined in Abaco. A total of 39 species of scleractinians and Millepora were observed in the Exumas (Chiappone et al., 1997) compared with a similar number of taxa (38) in Abaco (this study). The only previous examination of the reefs in the vicinity of the Hub of Abaco was carried out by Storr (1964) who examined the linear reefs offshore of the northern end of Elbow Cay between 1948 and 1961. His work was mostly qualitative in nature and his descriptions are of habitats that are landward of our outer reef crest site in Storr’ s Reef. Nevertheless, comparisons of general reef health between 1999 and the mid- twentieth century are interesting. Storr (1964) noted that Porites astreoides was abundant in all sites from offshore to inshore. Acropora palmata was abundant in the more seaward, shallow sites (reef flats), becoming patchy inshore. He described the appearance of A. palmata as healthy except at an inshore site where the corals were dead and covered with macroalgae. In contrast, A. palmata observed in this study showed high values of partial-and/or total- colony mortality but it is unknown when these colonies began to die. Porites astreoides, however, is still abundant. The most dramatic difference between the two time periods was the great abundance of Diadema antillarum found by Storr compared with the very low densities observed presently, a continuing result of the Caribbean-wide 1983 die-off (Lessios et al., 1984). Storr’ s (1964) observations of algal cover are too general for detailed comparison with our study (which otherwise might have allowed some correlation between abundance of Diadema and algal coverage). Comparison of the study sites within protected (from fishing) and unprotected areas showed slightly higher coral cover in the protected sites. Moreover, the largest massive corals were found on the Fowl Cay pinnacles, which are protected, and in the Lynyard Cay spurs and grooves, which are distant from human development. Interpretation of these differences should be made cautiously, however, since the protected areas were chosen for their high quality (as observed by divers) and there are 69 relatively few replicates (four protected versus nine that are unprotected). A number of confounding natural factors, such as depth, substratum, slope, structural complexity and exposure to wave energy can affect coral populations so further study is warranted before the effect on the coral community of protecting these Abaco reefs can be evaluated. The predominance of corals such as Porites astreoides, Diploria clivosa, Diploria strigosa and, in some cases, Acropora palmata, indicates that many of the surveyed sites are chronically exposed to relatively high wave energy and surge. The absence of large mounding corals, except in the deeper habitats on the Fowl Cay pinnacles (where small species, such as Agaricia spp., are also very common) and in the spur-and-groove formations at Lynyard Cay, is striking. The high relative abundance of Porites astreoides recruits is consistent with their high frequency as adult colonies in the surveyed reefs and their brooding mode of larval development (Chomesky and Peters, 1987). Due to the ambiguities of coral recruit taxonomic identification, the data are difficult to interpret with respect to relative abundance of species of Diploria, Montastraea, etc. which, when small, are only identifiable to the genus level. The relatively few recruits encountered (Table 4) also precluded interpretation of patterns of recruitment at the level of site or reef type. Many colonies of Montastraea faveolata and Diploria strigosa were observed with pale and dead tissue patches, perhaps due to environmental or biological stressors. Caribbean- wide warming occurred in the summer of 1999 and high-temperature exposure can result in bleaching and partial mortality (Williams and Bunkey Williams, 1990). Alternatively, coral diseases may have impacted these corals. It can be difficult to ascribe tissue necrosis to specific disease conditions (e.g., Richardson, 1998). Black-band disease has the clearest etiology; however, white-band disease, white plague, and other “white- line diseases” are harder to diagnose during snapshot field surveys. Some of the lowest (both Man O’ War Cay sites) and highest (Elbow Cay north) values for partial-colony mortality and for standing dead were found within the same reef habitat type (low-relief outer, reef crests). There were no apparent differences in terms of anthropogenic impact, either direct or indirect, in any of these reefs, nor were there clear natural impacts that could account for the differences. As few replicates were present for the other habitat types, it is not possible to generalize about coral mortality distribution patterns. Turf algae were very common but were trapping little sediment and appeared to have caused no stony coral mortality (personal observations). Macroalgal height and relative abundance trends may show some relationship to the degree of human impact in the region, as they are relatively low in the southern Sandy Cay fringing reefs, and some of the highest values occurred in the Elbow Cay reefs near Hopetown (Table 4). Insight into macroalgal biomass trends may be found by examining the role of herbivorous fishes in controlling macroalgae. Large groupers that were observed in the northern Fowl Cay pinnacles are protected from fishing and subjected to fish feeding by the local dive charter operators. Predation by carnivorous fishes off Fowl Cay may locally deplete the herbivorous fish populations leading to an increase in macroalgal biomass. Sandy Cay is similarly protected from fishing but not subjected to significant fish feeding and few large groupers were observed here, which may help to explain its lower macroalgal biomass values. 70 The AGRRA methods provide a good snapshot of the reefs of central Abaco and, as the results are compared to those of other Western Atlantic and Caribbean reefs (Kramer, this volume), may reveal large-scale trends in their stony coral, algal and fish communities. Management strategies on a local scale, however, should depend on studies that investigate the processes that affect reef communities. These investigations should include the impacts of anthropogenic factors such as water quality, fisheries and anchor damage, as well as the natural physical processes that may play an important role in controlling reef structure. ACKNOWLEDGMENTS This project was funded and assisted by the Caribbean Environment Programme of the United Nations Environment Programme, the Bahamas Reef Environmental Education Foundation, the National Coral Reef Institute at Nova Southeastern University Oceanographic Center and the Nova Southeastern University Farqhar Center for Undergraduate Studies. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. The Friends of the Environment, Abacos, Bahamas funded the use of one of our boats. Robert Ginsburg, Phil Kramer and Patricia Kramer were the driving forces behind this project, providing suggestions and encouragement at critical moments. Judy Lang offered many important suggestions that have improved the manuscript. Special thanks from the senior author and other participants are extended to Ken Banks who, in addition to assisting with this project, shared his house and boat with the team. REFERENCES Ambrose, H.W., and K.P. Ambrose 1995. A Handbook of Biological Investigation. Hunter Textbooks, Inc., Winston- Salem, N.C., 194 pp. Chiappone, M., K.M. Sullivan, and R. Sluka 1 997. Reef invertebrates of the Exuma Cays, Bahamas. Part I - Coral. Bahamas Journal of Science 4:30-36. Chomesky, E.A., and E.C. Peters 1987. Sexual reproduction and colony growth in the scleractinian coral Porites astreoides. Biological Bulletin 172:161-177. Dodge, S. 1999. The Cruising Guide to Abaco Bahamas 1999. White Sound Press, New Smyrna Beach, 136 pp. Glynn, P.W. 1973. Aspects of the ecology of coral reefs in the Western Atlantic region. Pp. 271-324. In: O.A. Jones and R. Endean (eds). Biology and Geology of Coral Reefs. II Biology 1. Academic Press, New York. 71 Goreau, T.F. 1959. The ecology of Jamaican coral reefs. 1. Species composition and zonation. Ecology 40:67-90, Humann, P. 1996. Reef coral identification: Florida, Caribbean, Bahamas. 3“^*^ Printing. New World Publications, Inc., Jacksonville, FL, 239 pp. Lessios, H.A., D.R. Robinson, and J.D. Cubit 1 984. Spread of Diadema mass mortality throughout the Caribbean. Science 226:335-337. Locker, S.D. 1 980. Origin and Depositional History of a Semi-enclosed Windward Lagoon off Great Abaco Island, Bahamas. University of North Carolina at Chapel Hill, Master’s Thesis, 181 pp. Richardson, L.L. 1998. Coral diseases: what is really known? TREE, 13:438-443. Storr, J.F. 1 964. Ecology and oceanography of the coral-reef tract, Abaco Island, Bahamas, Geological Society of America Special Paper Number 79, New York, 98 pp. Williams, E., and L, Bunkey-Williams 1990. The world-wide coral reef bleaching cycle and related sources of coral reef mortality. Atoll Research Bulletin 335:1-71. Woodley, J.D., P. Alcolado, T. Austin, J. Barnes, R. Claro-Madruga, G. Ebanks-Petrie, R. Estrada, F. Geraldes, A. Glasspool, F. Homer, B. Luckhurst, E. Phillips, D. Shim, R. Smith, K. Sullivan-Sealy, M. Vega, J. Ward, and J. Wiener 2000. Status of coral reefs in the northern Caribbean and western Atlantic. Pp. 261-285. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science. Cape Ferguson, Queensland and Dampier, Western Australia. Table 1. Site information for AGRRA stony coral and algal surveys in Abaco, Bahamas. 72 73 Table 2, Relative abundance of all stony corals seen during AGRRA surveys in Abaco. Name Relative abundance' Name Relative abundance' Milleporidae Millepora alcicornis Abundant Millepora complanata Abundant Scleractinia Acropora cervicornis Rare Montastraea annularis Abundant Acropora palmata Abundant Montastraea faveolata Abundant Agaricia agaricites f agaricites Abundant Montastraea franksi Rare Agaricia agaricites f danai Common Montastraea cavernosa Common Agaricia tenuifolia Few Manicina areolata Abundant Colpophyllia natans Common Meandrina meandrites Few Dendrogyra cylindrus Rare Muss a angulosa Few Dichocoenia stokesi Common Mycetophyllia danaana Few Diploria clivosa Abundant Mycetophyllia ferox Common Diploria labyrinthiformis Common Mycetophyllia Common lamarckiana Diploria strigosa Abundant Porites astreoides Abundant Eusmilia fastigiata Common Porites branneri Common Favia fragum Common Porites porites Common Isophyllastrea rigida Rare Scolymia spp. Common Isophyllia sinuosa Rare Stephanocoenia intersepta Common Leptoseris cucullata Common Siderastrea radians Common Madracis decactis Few Siderastrea siderea Abundant Madracis pharensis Few Tubastraea coccinea Rare 'Abundant = seen numerous times on more than half of the 13 dives, Common = seen on less than half of all dives but more than 10 times, Few = seen 4 to 10 times, Rare = seen 3 times or less. Relative abundance categories were determined during debriefings of all Abaco team members following each day of diving and at the end of the expedition. 74 o o o3 X) < _C (L» c/5 X5 (U +-* (U s 2 s o o c/) 03 O o >. c o o3 4-, O _o •I— > .2 ’> c/) +1 C 03 (U C O "O c o CJ -o c o3 OJ _N r<^ -2 Xi 03 H Table 4. Algal characteristics, and density of stony coral recruits and of Diadema antillarum (mean ± standard deviation) by site in Abaco. 75 76 Figure 1. Location of Andros Island, the Bahamas (inset) and the 1997 and 1998 AGRRA survey sites in four areas (North, Central, Bights, South) Andros. See Tables lA, IB for site codes. ASSESSMENT OF THE ANDROS ISLAND REEF SYSTEM, BAHAMAS (PART 1: STONY CORALS AND ALGAE) BY PHILIP A. KRAMER,' PATRICIA RICHARDS KRAMER,' and ROBERT N. GINSBURG' ABSTRACT Spatially extensive shallow reef crests with high coral cover (particularly Acropord) and fore reefs with high topographic complexity were documented during the 1997-1998 rapid assessments of coral and algal indicators in the relatively remote Andros reef system. Apart from the ecological effects of the Diadema antillarum die-off, evidence of major disturbance events (hurricanes, bleaching) were not apparent in 1997, although chronic levels of disease and predation were higher than in other Western Atlantic reef areas. High macroalgal abundances in fore reefs are attributed to low grazing pressures whereas the high densities of herbivorous fishes limit macroalgae in reef-crest habitats. During 1998, warm sea surface temperatures triggered localized bleaching in reef crests and appear to have facilitated widespread outbreaks of disease with resulting high mortality of massive corals in many fore reefs. INTRODUCTION Recent declines in abundance and diversity of corals in the Western Atlantic are well documented in areas where local stressors such as overfishing, nutrification and sedimentation can be significant (see Ginsburg and Glynn, 1994). Less well known are the condition of coral reefs in more remote areas where anthropogenic influences are less important but where regional impacts caused by bleaching, diseases and hurricanes may be significant. To better understand the influences of regional versus local impacts, we examined the condition of a relatively remote area in the Bahamas, the Andros reef tract. Andros is separated both geographically and hydrographically from the greater Caribbean which may decrease the exchange of larvae and/or water-borne pathogens with other reefs of the region. Regional disturbances, such as the die-off of many acroporids from white-band disease (WBD) (Aronson and Precht, 1997, 2001) and detailed effects of the 1983 Caribbean-wide Diadema urchin die-off (Lessios et ah, 1984), are poorly documented in Andros. ' Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: pkramer@rsmas.miami.edu 78 The extent that anthropogenic impacts have affected the reefs off Andros is poorly quantified but is generally assumed to be minimal compared with many areas of the Caribbean. Andros has a very low population of about 8,155 (Bahamas Reef Environmental Education Foundation and MacAlister Elliot and Partners, 1998, unpublished report) with most of its populace concentrated in several small towns along the eastern coast of the island. Androsians rely on fishing, sponging and small-scale tourism for the majority of their income while additional employment comes from the Atlantic Underwater Test and Evaluation Centers (AUTEC), Androsia (a local batik factory), and small-scale farming. North Andros, used for timbering until 1975, now contains a small-scale agricultural industry (Sealey, 1990) but pollution and runoff are believed to be limited. The few small-scale resorts on Andros (<300 rooms total) cater towards tourism focused on guided bonefishing or diving the coral reefs. Spatial trends in various indicators of the principal reef-building corals and algal populations that were surveyed during 1997 and 1998 using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) method are described in this paper. The condition of reef fish populations in Andros is presented in Kramer, Marks and Turnbull (this volume). METHODS Andros Island has the largest land area (6,000 km^) of the Bahamas Archipelago and is located on the eastern margin of the western arm of Great Bahamas Bank where it abuts against the Tongue of the Oeean (Fig. 1). The extensive Andros reef tract parallels the eastern side of the island from Joulter Cays in the north to Saddleback Cays in the south (a total of 217 km). The reef is discontinuous in many places, particularly around the eentral and southern portion of Andros where several large channels (bights) cut across the entire island dividing it into three distinct landmasses. Smith (1948) and later Newell et al. (1951) both used the term "bank reef to describe the Andros reef tract although it has also been described as a fringing reef, barrier reef, and fringing-barrier reef (which is probably the most accurate). The reef crest typically lies 1-2 km from the shoreline and is situated atop a Pleistocene ridge that is exposed in places in the form of small elongate offshore cays (e.g., Staniard Rock, Goat Cay, Long Rock, Long Bay Cays, Grassy Creek Cays). Fore reefs are best developed as high-relief features (e.g., pinnacles) composed of often dense aggregations of massive corals (primarily the Montastraea annularis species complex) and occur at intermediate depths between 7 m to 12 m, being most extensive in “low-to-intermediate” wave-energy environments. Well-developed reef crests and fore reefs often occur together in central and southern Andros. The outer reef slope contains a break at approximately 20 m followed by the shelf edge at 30 m that is marked by a near-vertical wall rich in sponges and other invertebrates. For this study, the Andros reef tract was stratified into shallow crests (1-3 m deep) and intermediate-depth fore reefs (8-12 m deep), which were further subdivided into “well developed” and “poorly developed” categories for both depths using low altitude (200 m) oblique aerial photographs taken in July, 1997. Lagoonal patch reefs and deeper shelf-margin reefs were not investigated. Recormaissance inspections or Manta tows were undertaken in each area to get an overview of reef type before haphazardly selecting representative well-developed sites to survey. In 1997, 26 sites (13 at 1-3 m depth, 13 at 79 8-12 m) located primarily in the northern and southern sections of Andros Island were examined. In 1998, 28 (15 deep and 13 shallow) sites along 75 km of central Andros were surveyed (Fig. 1; Tables lA, IB). Four sites overlapped between the 1997 and 1998 surveys (S6-S7; D6-D7; D9-D10; D12-D13). All surveyors were trained in the AGRRA method and consistency exercises were conducted regularly to minimize differences among observers. For the two survey periods, a total of six persons participated in the benthic surveys with three surveyors collecting approximately 75% of all the data. The total number of 1 0-m benthic transects deployed at each site ranged from 6 to 24 and video was recorded on some transects for archival purposes. In 1997, when a modified Version 1 of the AGRRA protocols was used, live stony coral cover was not quantified. Visual estimates of old mortality did not factor out open spaces between columns for lobate colonies such as M annularis and recent mortality estimates included sides of colonies. Maximum (rather than average) macroalgal heights were recorded within quadrats. In addition, after estimating their abundance in the uppermost canopy layer of the algal quadrats, macroalgae were removed prior to estimating the abundance of underlying turf and crustose coralline algae. We substituted the newly adopted Version 2 of the AGRRA protocols for the 1998 surveys (see Appendix One, this volume), expanding the size and health assessments to all stony corals that were at least 10 cm in diameter for approximately half of the sites. Coral sizes were measured to the nearest 5 cm in both years. Identification of corals was based on Humann (1992). Statistical analysis was performed with the Statistica (version 5.1) program. To permit comparisons between years, all stony corals <25 cm in diameter were excluded from these analyses and the Montastraea annularis species complex was treated as a single entity. To statistically characterize variance for the benthic indicators, only the first 10 transects were analyzed from each site. Size data were log(X+l)-transformed and percentage data were arcsine-transformed. Parameters were analyzed by students t-test and by 1-way and 2-way Analysis of Variance (ANOVA). For spatial analysis, the Andros sites were divided into four areas (North, Central, Bights, and South) (Fig. 1) and analyzed by ANOVA with sites hierarchically nested under areas as random factors. RESULTS Species Composition and Abundance of Stony Corals A total of 3,221 stony corals were examined in 28 reef crests in Andros during the two years of field work. They were dominated by colonies of Acropora palmata and displayed varying degrees of development controlled, in part, by wave energy, reef aspect, and the presence of freshwater creeks. China Point (SIO) was one of the few shallow sites that contained monotypic thickets of living A. cervicornis, although “standing dead” (colonies completely dead and still in growth position) thickets were encountered at several other sites. Live stony coral cover ranged from 20-54% (mean=36, sd=18) in 1998 in shallow reefs (Table lA). A total of 16 “large” (>25 cm) scleractinian and one hydrozoan species were recorded in the shallow crests (all reefs combined). Acropora palmata comprised a mean of 62% of all these large corals (Fig. 2A). The relative abundance of subdominant taxa on reef crests (e.g., M. annularis, Millepora complanata, Pontes astreoides) was remarkably uniform across the reef areas (Fig. 3A). The most common size range for the colonies of A. palmata was 120-140 cm (mean diameter=137, sd=77) with the largest colonies exceeding 400 cm in diameter (Fig. 4A). p/ 4"/ A B Taxon codes AC = Acropora cervicornis AP = .4. palmata AG = Agaricia spp. CN = Colpophyllia natans DS = Diploha strigosa DL = D. labyrinthiformis MA = Montastraea annularis MAP = M. faveolala MC = M cavernosa MFR = M franksi MIL = Millepora spp. PA = Porites astreoides PP = Porites porites SS = Siderastrea siderea N = 3233 N = 3198 Figure 2. Species composition and mean relative abundance of all stony corals (>25 cm diameter) on (A) reef crests and (B) fore reefs in Andros. Its colony diameters were significantly smaller in the bights area (mean=l 18 cm for S18- S22) compared with the other areas along Andros (mean=152 cm) (ANOVA df=l, MS=37, F=18.2, p<0001) (Fig. 4A). Average colony diameters overall were largest in the northern area and smallest in the bights area (Table 2A). In the 32 intermediate-depth fore reefs that were surveyed, a total of 3,156 large corals were assessed. Fore reefs were mainly dominated by assemblages of M. annularis (Fig. 2B) and they were best developed around protected channels, cays, and seaward of reef crests. Live stony coral cover in 1998 ranged from 6 to 45% (mean=22.6, sd=13.8) in the fore reefs (Table IB). Several sites in the bights area (e.g., D18, D19) had low coral cover (<10%) and one site (D23) contained few corals greater than 25cm in diameter. A total of 1 8 species of large stony corals were counted within fore reef transects (all reefs combined), 67% of which were M. annularis complex (M. annularis, M. faveolata, M. franksi) (Fig. 2B). An exception to this pattern was the dominance of S. siderea rather than M. annularis at several sites in the bights area (Fig. 3B). Colonies of M. annularis were most common in the 40-50 cm size range (average=64, sd= 8), yet some individuals exceeded 250 cm in diameter (Fig 4B). Colony sizes were negatively correlated with water depth (r =0.03, p<0.0001) and significantly smaller in the reefs surveyed in 1998 (mean diameter=54 cm) than in 1997 (mean diameter=70 cm) (t-test, df=1443, t=10.1, p<0.0001). As in the reef crests, colonies were largest in the northern area and smallest in the bights (Table 2A). The number of species of stony coral recruits (^2 cm) was greater in fore reefs (26 species, all reefs combined) than in reef crests (16 species, all reefs combined). Small stony coral densities were also significantly higher in deeper (equivalent to 1 0.6 recruits/m^) than in shallower sites (equivalent to 2.9 recruits/m^) (t-test, t=12.7, df =544, 81 12 B □ North (n= 739) □ Middle (n= 963) Bights (n=438) South (n=l 058) Taxon codes AC = Acropora cervicornis AGA = Agaricia spp. CN= Colpophyllia natans DC = Diploria clivosa DL = D. labyrinthiformis DS = Diploria sirigosa MC = M cavernosa MILC = Millepora complanta PA = Porites astreoides PP = P. porites SS = Siderastrea siderea MILC MC AC PA PP AGA DS DC DL SS Figure 3. Relative species abundance of the 10 most abundant stony corals (>25 cm diameter) on (A) reef crests (here excluding Acropora palmata), (B) fore reefs (here excluding the Montastraea annularis complex) in four areas of Andros. p<0.0001) but showed a high degree of variability among reefs (Tables 3A,3B). Porites astreoides was the most abundant recruit in reef crests (50%) and fore reefs (25%) despite low abundances of large adults (~4% in reef crests, 3% on fore reefs) (Fig. 5). Acropora palmata was the third most abundant (9%) recruit in shallow reefs with the M. annularis complex as the fourth most abundant (6%) in deep reefs. 82 A A. palmata B M annularis O a 25 cm Figure 4. Size-frequency distribution of all colonies (>5 cm diameter) of (A) Acropora palmata on reef crests and (B) Montastraea annularis on fore reefs in four areas off Andros. Stony Coral Condition Levels of old partial-colony mortality (hereafter old mortality) in large stony corals ranged from 12% (S5) to 52% (S25) (Table 2) and were higher in reef crests than fore reefs for both survey years (2-way ANOVA, df=581, MS=1787, F=12.5, p<0.0001). The frequency of 100% standing dead colonies was 10%. Reefs surveyed in 1997 had significantly higher old mortality (crests = 38%, fore reefs = 29%) than those surveyed in 1998 (crests = 29%, fore reefs = 22%) (2-way ANOVA, df=581, MS=5665, F=40.4, p<0.0001). Large differences in old mortality were apparent among species, with acroporids (n=l,764) accounting for the highest levels of old mortality (48% for^. 83 N Taxon codes AG = Agaricia spp. AP = Acropora palmata CN = Colpophyllia natans Die = Dichocoenia stokesi DS = Diploria strigosa FF = Favia fragum MA = Montastraea "annularis ” MC = M. cavernosa MIL = Millepora spp. MY = Mycetophyllia spp. PA = Porites astreoides PP = P. porites SC = Scolymia spp. SS = Siderastrea siderea UK = Unknown Figure 5. Species composition and mean relative abundance of stony coral recruits (<2 cm diameter) in (A) reef crests and (B) fore reefs off Andros. Figure 6. Percent old and recent (rec) partial colony mortality (mean ± standard error) in 1998 for the 15 most abundant >25 cm stony coral species; both survey depths combined, off Andros. Species codes same as in Figure 5 except MAF = Montastraea faveolatz., MA = M. annularis, MFR = M.franksx, AC = Acropora cervicornis, DL = Diploria labyrinthiformis, MILC = Millepora complanata, DC = Diploria clivosa. cervicornis, 38% for^^. palmata). These high values can be largely attributed to the presence of 100% standing dead colonies oi Acropora which made up 1-24% (average 7%) of the assessed corals in shallow reef crests. Excluding 100% standing dead colonies, Acropora had an average old partial-mortality of 21%. Old mortality at fore reef sites ranged from 13% (D15) to 45% (D32) and the frequency of 100% standing A. palmata freef crests^ Colony diameler Colony diameler (cm) 40 , ^ ^ ^ ^ . — C Standing dead 30 n= 123 corals Colony diameter (cm) Figure 7A. Size frequency distribution of colony diameter for Acropora palmata for 1998 with (A) no, (B) partial (1-99%) or (C) 100% (=standing dead) old colony mortality in reef-crest sites off Andros. 85 M annularis (Tore reefs^ Colony diameter (cm) Colony diameter (cm) C Standing dead n=30 R5SS8R5S88 ^^S^ocbocicb II O fN ID CD A Colony diameter (cm) Figure 7B. Size frequency distribution of colony diameter for Montastraea annularis for 1998 with (A) no, (B) partial (1-99%) or (C) 100% (= standing dead) old colony mortality in fore-reef sites off Andros. 86 dead colonies averaged 7% (Table 2B). Montastraea faveolata and A. palmata had the highest mean old mortality values (>30%) while Pontes astreoides had the lowest (<10%) (Fig. 6 for 1998 data). In both A. palmata and M. annularis, old mortality values increased with colony size up to the mode of the population (120-140 cm-size classes for reef-crest colonies of A. palmata, 40-50 cm size classes for fore-reef colonies of M annularis). Smaller corals tended to be either completely alive or completely dead (i.e., to have either no mortality or 1 00% mortality) whereas areas of old mortality were displayed on the surfaces of many of the larger colonies (Fig. 7A, B). Examination of three spatial scales (within reef, within areas, between areas) for each reef type showed that within-reef-variation explained approximately 80% of the variation in old mortality. Recent partial-colony mortality (hereafter recent mortality) in the >25 cm stony corals ranged from 0.5-42.5% (Table 2) and was higher in fore reefs (mean 13%) than in reef crests (mean 6%) during both survey years (2-way ANOVA, df=580, MS==3403, F=34.6, p<0.0001). Recent mortality in the reef crests was lower during the 1998 surveys (average 4%) than in 1997 (average 7%) (t-test, df=270, t=5, p<0.0001) but was not significantly different between years in the fore reefs (1997 mean=l 1%, 1998 mean=13%) (t-test, df =310, t=0.4, p=0.7). Recent mortality showed no significant correlation with colony diameter in A. palmata and M annularis. Very high levels of recent mortality (>20%) were recorded at several fore reefs in central and southern Andros during the 1998 survey (DIO, D 13, D24, D25). Significant differences in recent mortality also occurred between different coral species in 1998 (1-way ANOVA, df =40, MS=3787, F=7.7, p<0.001). Corals with the highest levels of recent mortality included Montastraea franksi, Colpophyllia natans, M. annularis, and Diploria labyrinthiformis (Fig. 6 for 1998 data). Identifiable causes of recent stony coral mortality in reef crests included predation, diseases (Table 2A), physical damage associated with human-induced trash and boat groundings, and a bleaching event associated with warmer seawater temperatures observed in summer 1998 (see Kramer, this volume). Many colonies of^. palmata and A. cervicornis had elevated “chimneys” characteristic of repeated predation by the three-spot damselfish, Stegastes planifrons. Damselfish gardens were present on approximately 8% of all the surveyed stony corals and were significantly more common in reef crests than in fore reefs (t-test, df=584, t=3.76, p<0.0001). Evidence of predation by parrotfishes (particularly the stoplight parrotfish, Sparisoma viride) in the form of excavations or small bite marks was low. The corallivorous snail, Coralliophila abbreviata, was observed singularly and in small groups (three to five per group), mainly on colonies of Acropora. Small white (<25 cm^) lesions were present on approximately 20% of all A. palmata surveyed (both years combined) and were attributed to either disease (patchy necrosis = white pox) or predation by Coralliophila. Although distinguishing recent mortality due to patchy necrosis from predation was often difficult, on average about 9% of the colonies of^. palmata were affected by diseases (all shallow sites combined). High (12-15% in S7, S9, S13, S15, S26) to alarmingly high (-25-30% in SI 1, SI 2, S28) incidences of disease were also observed. No large-scale bleaching was observed during 1997 (<2% of colonies overall affected); however, during August 1998, substantial bleaching (>10% of colonies affected) was observed in several shallow sites (e.g., S5, S21). Bleaching was very species-specific with Millepora complanata, Agaricia tenuifolia, A. palmata, and yl. cervicornis being 87 most affected. At one site (S24), where 80% of the colonies were discolored (mostly being completely bleached), entire “very recent” colony mortality was observed in several instances. Recent mortality of stony corals in the fore reefs was mainly attributed to diseases (Table 2B) and, to a lesser degree, to overgrowth by macroalgae. No significant predation or bleaching was noted either year in deeper water (> 5m). In 1997, macroalgal overgrowth by Microdictyon marinum was observed to cause limited bleaching and mortality around the edges of colonies, particularly the lobate M. annularis. Stony coral diseases were widespread during 1997, but at moderate levels (4% of colonies infected). During 1998, significantly higher levels of disease were recorded (18% infected colonies) (1-way ANOVA, df =1, MS=1418, F=21.7, p<0.001), with the most common being black-band disease (BBD) and white plague (WP). During 1998, active BBD and WP infected mainly faviid corals which included in order of decreasing infection M. franksi, M. annularis, M. faveolata, M. cavernosa, D. labyrinthiformis and C. natans. Within reefs, BBD occurred in clumped distributions with simultaneously active infections in colonies that were either touching or closely adjacent. At larger spatial scales, diseases were more prevalent in fore reefs with higher colony densities than in less well- developed reefs with lower colony densities. This pattern may be due in part to the abundance of M. franksi and M annularis which were more susceptible. The incidence of WP varied among reefs and it was observed most commonly in the central and southern areas in reefs containing a high density of large corals. Here, sites with <5colonies/10m often had a lower prevalence of disease. Both BBD and WP affected all stony coral size classes. Approximately 6% of all fore-reef stony corals surveyed in 1998 (primarily M. annularis and M. franksi) had suffered 100% recent mortality. The average colony diameter of the >25 cm corals with 100% recent mortality was 44 cm with most being in the 25-40 cm size class (although corals <25cm were not surveyed in all sites). During 1998, the percentage of recent partial-colony mortality was strongly correlated with the percentage of stony corals showing signs of disease (r^=0.63, p <0.001). Algal Communities and Diadema antillarum Algal populations off Andros were distinctly different between the two reef habitats but were remarkably consistent within each depth zone. Reef crests (Table 3A) tended to be dominated by crustose coralline algae (mean relative abundance = 44%) and turf algae and contained relatively few macroalgae (mean relative abundance =15%). Dominant macroalgal species in the reef crests were Dictyota divaricata, Stypopodium zonale and Turbinaria turbinata with average canopy heights from 1-3 cm (mean = 1.8 cm) in 1998. In contrast, macroalgae (mean relative abundance = 47%) and crustose corallines (mean relative abundance = 35%) were codominants in fore reefs where turf algae were relatively scarce (Table 3B). The principal macroalgae observed in fore reefs were Microdictyon marinum, Lobophora variegata, Dictyota pulchella, Dictyota divaricata, Sargassum hystrix, and Halimeda spp. Macroalgal heights were significantly higher in fore reefs (averaging 3.3 cm in 1998) than in reef crests for both survey years (2-way ANOVA, MS=2.9, df =534, F=97, p<0.0001). When macroalgal abundance and height are considered together as a macroalgal index (Tables 3A, 3B), the index was four times greater in fore reefs (mean=204) than in reef crests (mean=46). Diadema was not 88 recorded in any of the sites during 1 997 although a few individuals were counted outside of transects at several sites. Diadema densities were slightly higher at examined sites in 1998, averaging 0.4 individuals/lOOm^ in both habitats. DISCUSSION Shallow Reef Crests The reef crests off Andros presently contain extensive areas of Acropora palmata, a large percentage of which have colonies that are partially or even fully alive on their upward facing surfaces. Densities of live and partially live (i.e., <100% total partial- colony mortality) colonies were high (~4 colonies/1 Om). Evidence of the WB epizootic reported to have decimated populations of A. palmata populations in other areas of the Caribbean (Aronson and Precht, 1997, 2001) was minimal. The low-to-moderate levels of old mortality, large colony sizes, and fairly even size-distributions that we observed all suggest modest levels of disturbance to A. palmata in the last several decades. Smaller colony sizes were recorded in several sites, particularly near the bights area (e.g., S16- S22) where tidal creeks and channels discharge water having highly variable temperatures and salinities from Great Bahama Bank that probably inhibits Acropora growth. When cervicornis was present on shallow reefs, it usually occurred as thickets, particularly in the back reef, although most of these colonies were long dead. The distribution and abundance of historic A. cervicornis populations are not known for Andros although anecdotal reports suggest it may have been more common in shallow- patch and back-reef habitats. Low-to-moderate values (1-5%) for recent partial-colony mortality observed in the majority of the reef crest surveys is thought to represent background (chronic) mortality caused by predation, disease, competition, and other biotic interactions. The higher levels of recent mortality recorded during 1997 (e.g., S9, SI 2- 14) are mainly attributed to the way mortality was scored (by including colony sides) rather than to higher levels of disturbance relative to 1998. Levels of recent mortality recorded during 1998 at non- bleached sites (~3.5%) are comparable to levels recorded in other western Atlantic reef crests not experiencing acute disturbances (Kramer, this volume). Most colonies of A. palmata had recently dead tissues; however, the small size of most of these lesions suggests either that routine injuries are of limited areal extent and/or tissue regeneration occurs fairly quickly (Meesters et al., 1996). Alternatively, if lesions were caused by disease, tissue loss is slow but may accumulate while the colony remains infected. National Oceanographic and Atmospheric Administration (NOAA) hotspot maps indicated a 1 -2 degree increase in sea surface temperatures beginning in late May, 1 998 and extending through August, 1998. Observations during 1998 suggest that bleaching and bleaching-induced mortality off Andros mainly impacted the shallow reefs. Spatially, the extent of bleaching impacts to shallow reefs was variable with reefs immediately south of South Bight being the most affected during the time of our survey. The high value for recent mortality (>15%) observed in one reef crest (S24) surveyed during August, 1998 is attributed to bleaching-induced mortality. Species that suffered mortality 89 were mainly Millepora complanata, Agaricia tenuifolia and, to a lesser extent, A. palmata. Local reports from the Bahamas during 1998 suggest that region- wide bleaching reached a peak in late August/early September at least two weeks after our surveys ended (Wilkinson, 1998). Therefore, while our surveys indicated only localized impacts, the full effects of the 1998 bleaching event are difficult to ascertain since the immediate damage may have ensued for several weeks to months after our surveys and occurred beyond our survey area. However, the low relative abundance of macroalgae on reef crests, high abundance of living^, palmata colonies, and the presence of A. palmata recruits (mean=0.27/m ) suggest that recovery from low-to-moderate levels of bleaching-induced mortality on reef crests is favored. Fore Reefs Patterns in coral abundance in the Andros fore reefs were evident at both small and large spatial scales. Well-developed fore reefs at intermediate depths (7-12 m) occurred mainly as dense patches of the M annularis species complex often located off islands and along channels where either protection from large swells and/or underlying Pleistocene geomorphology favored reef growth. Grading away from these dense patches were stretches with lower relief, lower coral densities, and smaller coral sizes. Hence, the degree to which fore reefs had developed was largely determined by the abundance of the M. annularis complex. In poorly developed reefs, such as N. Mangrove (D23), the dominant coral was S. siderea (44%) while the M. annularis complex comprised only 1 7% of the large (>25 cm in diameter) coral population. The size and abundance of the colonies of Montastraea in any particular site presumably reflect both the suitability of local conditions (Bak and Meesters, 1998) and rates of coral growth. That fore reefs located in shallower water (7-9 m) had larger colonies of the M annularis complex than those located slightly deeper (10- 12m) is probably explicable by the observation that growth rates for Montastraea can decrease sharply in this depth range (Hubbard and Scaturo, 1985) and even small absolute increases of water depth (2-3 m) may shift colonies to smaller sizes. The smaller colony sizes found overall in the 1998 surveys can be mainly attributed to the relatively poor environmental conditions in many of the reefs near the bights. In addition, survey depths were slightly shallower in 1997 compared with those in 1998 because the zone of maximum development in the northern and southern areas was shallower than in the central and bights areas. That we found a strong relationship between levels of old mortality in M. annularis and colony size up to the mode of the population (40-50 cm) in Andros (Fig. 7) is similar to the relationship described by Bak and Meesters (1998) in the Netherlands Antilles. In addition, levels of old mortality were strongly influenced by the species composition of the stony corals present in any given reef. Therefore, differences in mean colony size and species composition can explain some of the between-reef variability in levels of old mortality and these factors should be taken into account when comparisons are made among sites, reefs, areas, subregions, and regions. During 1998, a mass mortality event affected many massive corals in the Andros fore reefs. This widespread disturbance is thought to have begun in June approximately one month after abnormally warm sea surface temperatures enveloped the area (T. Turnbull, personal communication). Local observations of “white coral” were first 90 reported in early July. One month later we found many massive colonies with either partial or complete recent tissue mortality and signs of disease (primarily WP and BBD) but surprisingly little evidence of bleaching (e.g. <5% colonies displayed discolored tissues). At severely impacted fore reefs, we estimated that up to 50% of the live stony coral cover had been recently lost. We believe severe outbreaks of one or more diseases must have occurred in late June/early July and moved through the entire area at a very rapid pace, perhaps in response to the increased seawater temperatures. [Increased temperatures can trigger or exacerbate the activity of microorganisms (bacteria, cyanobacteria, fungi, protozoans) and possibly viruses that are responsible for outbreaks of disease (e.g., Rutzler et al., 1983).] That the highest mortalities off Andros were observed at depths of 6-12 m in reefs with high coral densities and complex geomorphological structure suggests that diseases might be expected to spread most rapidly under conditions in which colonies (particularly monogeneric coral assemblages) have direct or close contact with one another. The unusually luxuriant growth of the M. annularis species complex that is typical of the Andros fore reefs may have made these communities very prone to outbreaks of disease. Recovery from this massive disturbance will take many years, possibly decades. Acropora cernivornis was rarely seen in the deeper sites. Examination of sediments from several fore reefs, however, revealed only small amounts of A. cervicornis rubble indicating that it may never have been as common on Andros as in many other areas of the western Atlantic (e.g., Jamaica, Belize, Bonaire). Algae and Diadema Historic populations of Diadema in the Bahamas, particularly at shallow depths, were very high according to descriptions by Storr (1964) and Newell and Rigby (1951). The low abundances observed in Andros (both shallow and deep sites) during our surveys reinforces local anecdotal accounts that the 1983-84 die-off of Diadema antillarum in the western Atlantic (Lessios et al., 1984) also decimated populations on Andros. At present, scarids (parrotfish), acanthurids (surgeonfish), and some pomacentrids (damselfish) are the principal herbivores in Andros’ reefs (Kramer, Marks and Turnbull, this volume). As documented in other studies (Hay, 1984; Morrison, 1988), these herbivores are disproportionately distributed across the Andros shelf with higher numbers principally associated with well-developed shallow reef-crest habitats. Macroalgal indices for the 1998 sites displayed a strong inverse relationship with herbivore biomass (Kramer, Marks and Turnbull, this volume). The relationship is unusually well pronounced, and may also reflect that our surveys took place when, due to seasonal blooms of Microdictyon marinum, the macroalgal index reaches a seasonal peak in the late summer (T. Turnbull, personal communication). Within individual habitat types (e.g., reef crest or fore reel), the herbivore-algae relationship was only weakly significant indicating other synergistic factors, such as wave energy and sedimentation, may also influence algal abundance patterns. However, the overall algal-herbivore relationship between depths suggests a strong top-down control on the distribution of macroalgae on Andros. Herbivorous fishes may have compensated for the loss of Diadema as an herbivore in reef crests but not in fore-reef sites. High macroalgal canopy heights and lack of grazed surfaces reflect a general lack of herbivory at greater (>6 m) 91 depths. The impact of excessive algae to fore reef corals is potentially severe, but not well characterized, off Andros. Microdictyon, which attaches to the sides of Montastraea annularis columns, was observed to grow over the edges of its live tissues and other small corals causing bleaching and, in some cases, mortality. Following the 1998 mass mortality of many massive stony corals, macroalgal abundances are expected to have further increased on fore reefs given the increased availability of substratum space. Macroalgal abundances will likely remain high in Andros fore reefs until levels of grazing by either Diadema or herbivorous fishes increase. Even given the high relative abundance of macroalgae, mean densities of coral recruits in the Andros fore reefs were fairly high (equivalent to 10.6/m^) compared with other areas of the wider Caribbean (Bak and Engel, 1979), suggesting that small corals may be able to survive despite overtopping by macroalgae. In addition, average recruit density for the shallow sites (equivalent to 3 juveniles/m^) was higher than at similar reef types in the Florida Reef Tract (maximum ~2.5 juveniles of <4cm diameter/m^) (Chiappone and Sullivan, 1996). Juvenile species richness was also higher in the shallow reefs off Andros (16 species) than in shallow reefs off the Florida Keys (~1 1 species) (Chiappone and Sullivan, 1996). The presence and relatively high abundance of recruits of A. palmata in reef crests and of the M annularis complex in fore reefs is surprising given that recruits of these species are generally absent or present in very low numbers in settlement studies (e.g., Bak and Engel, 1979; Hughes and Jackson, 1985; Rylaarsdam, 1983). High recruit abundance for these corals on Andros may be attributed to the high densities of adult colonies. The survivorship of the recruits we surveyed was not examined but may play an important role in the stability of these populations. Additional surveys focused on understanding the recruitment and survivorship of A. palmata and the M annularis species complex on Andros are warranted. Despite their relatively remote location, our results indicate that the reefs off Andros Island have been affected by some (e.g., loss of Diadema antillarum), but not many, of the regional disturbances documented in other parts of the Caribbean. Impacts of the regional WBD epizootic on Acropora and of the massive bleaching events prior to 1998 appear to have been lower off Andros, based on the dense thickets of live A. palmata and the large size of many stony corals at the time of our survey. However, a high abundance of macroalgae and the presence of coral diseases in these reefs are suspected in years prior to 1997. In 1998, the massive disease outbreak in fore-reef habitats, particularly in the central and bights areas, resulted in catastrophic tissue loss for many large corals (particularly the M. annularis complex). Bleaching in shallow reef crests during 1998 also caused acute localized loss of live stony corals often resulting in complete colony mortality for the most severely impacted reefs in southern Andros (personal observations). The longer-term trajectory with regard to the recovery of these reefs will depend on the frequency of future disturbance events, the status of their Diadema populations, and the recruitment of new, and survivorship of existing, stony corals. 92 ACKNOWLEDGMENTS This project was made possible by a grant from the Committee for Research and Exploration of the National Geographic Society (Grant #6046-97). The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. Additional sponsorship was provided by the Hachette Filipacchi Foundation and Carolina Skiffs. Assistance with the field surveys was provided by V. Kosmynin, K. Marks, J. Renter, B. Riegl, E. Phillips, S. Lutz, and C. Peters. Special thanks are extended to W. Bryant, M. McGarry, and S. Cawthon for their adept piloting skills. The late Captain R. Gaensslen generously provided the use of his vessel. The Captain ’s Lady, for the 1997 survey. For the 1998 survey. Go Native Yacht Charters is thanked for partial charter sponsorship of their 60' catamaran. The Liahona. E. Fisher assisted with data entry and J. Lang provided many useful comments on an earlier version of this manuscript. We are grateful to the Bahamas Department of Fisheries for granting permission to conduct field work in Andros. REFERENCES Aronson, R.B., and W.F. Precht 1997. Stasis, biological disturbance, and community structure of a Holocene coral reef. Paleobiology 23:326-346. Aronson, R.B., and W.F. Precht 2001. Evolutionary paleoecology of Caribbean coral reefs. Pp. 171-233. In: W.D. Allmon and D.J. Bottjer (eds.). Evolutionary Paleoecology: The Ecological Context of Macroevolutionary Change. Columbia University Press, New York. Bak, R.P.M., and M.S. Engel 1979. Distribution, abundance and survival of juvenile hermatypic corals {Scleractinia) and the importance of life history strategies in the parent community. Marine Biology 54:341-352. Bak, R.P.M., and E.H. Meesters 1998. Coral population structure: The hidden information of colony size-frequency distributions. Marine Ecology Progress Series 162:301-306. Chiappone, M., and K.M. Sullivan 1996. Distribution, abundance and species composition of juvenile scleractinian corals in the Florida Reef Tract. Bulletin of Marine Science 58:555-569. Ginsburg, R.N., and P.W. Glynn 1994. Summary of the colloquium and forum. Pp. i-ix. In: R.N. Ginsburg (compiler). Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards and History, 1993. Rosenstiel School of Marine and Atmospheric Science. Miami. FL. Hay, M.E. 1984. Patterns of fish and urchin grazing on Caribbean coral reefs: are previous results typical? Ecology 65:446-454. 93 Hubbard, D.K., and D. Scaturo 1985. Growth rates of seven species of scleractinian corals from Cane Bay and Salt River, St. Croix, USVI. Bulletin of Marine Science 36:325-38. Hughes, T.P., and J.B.C. Jackson 1985. Population dynamics and life histories of foliaceous corals. Ecological Monographs 55:141-166. Humann, P. 1992. Coral Reef Identification, Florida, Caribbean, Bahamas. New World Publication, Inc. 278 pp. Lessios, H.A., D.R. Robertson, and J.D. Cubit 1984. Spread of Diadema mass mortality through the Caribbean. Science 35 335-337. Meesters, E.H., 1. Wesseling, and R.P.M. Bak 1 996. Partial mortality in three species of reef-building corals and the relation with colony morphology. Bulletin of Marine Science 58:838-52. Morrison, D. 1998. Comparing fish and urchin grazing in shallow and deeper coral reef algal communities. Ecology 69:1367-1382. Newell, N.D., and J.K. Rigby 1951. Shoal-water geology and environs, eastern Andros Island, Bahamas. Bulletin of the American Museum of Natural History 97:1-30. Rylaarsdam, K.M. 1983. Life histories and abundance patterns of colonial corals on Jamaican reefs. Marine Ecology Progress Series 13:249-260. Rutzler, K., D.L. Santavy, and A. Antonius 1983. The black band disease of Atlantic reef corals. III. Distribution, ecology, development. Marine Ecology 4:329-358. Sealy, N.E. 1990. The Bahamas Today: An introduction to the human and economic geography of the Bahamas. MacMillan Caribbean. Hong Kong. 120 pp. Storr, J.F. 1955. Ecology and Oceanography of the Coral Reef Tract, Abaco Island, Bahamas. Ph.D dissertation. Cornell University, Ithaca, NY. 264 pp. Wilkinson, C. 1998. The 1997-1998 mass bleaching event around the world. Pp. 15-38. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 1998, Australian Institute of Marine Science, Cape Ferguson, Queensland. 94 Table 1 A. Reef crest site information for AGRRA stony coral and algal surveys off Andros Island, Bahamas {1997 sites are italicized). Reef crest Site Name Site Code Latitude (N°') Longitude (W“') Survey Date Depth (m) Transects (#) >25 cm Stony Corals (#/10mf % Live Stony Coral Cover (mean ± sd)^ N. Joullers SI 25.31322 78.03433 Aug 15 97 2.5 13 13.5 — Golding S2 25.22392 78.08557 Aug 18 97 1.5 12 12.0 ... Morgan S3 25.1601 78.0029 Aug 17 97 1.5 14 10.5 ... Coconut Point S4 25.12885 77.98303 Aug 17 97 1.5 13 11.0 ... Mahore S5 25.06367 77.93783 Aug 21 98 1.0 10 9.5 43.0 ± 14.5 S. Staniard 2 S7 24.84493 77.86098 Aug 20 98 0.5 7 9.0 41.0±22.5 S. Staniard 1 North Andros- S6 24.84217 77.8586 Oct 19 97 3.5 9 9.0 ... all' 78 10.6 42 N. Love Hill S8 24.77435 77.80772 Aug 18 98 1.0 9 9.5 41.5 ± 14.0 5’, Love Hill S9 24.84425 77.8563 Aug 5 97 1.5 17 7.0 ... China Point SIO 24.75133 77.80767 Aug 18 98 1.0 13 9.0 54.0±3L0 Red Rock Sll 24.72917 77.77017 Aug 7 98 1.0 17 6.5 37.0± 15.5 S. Autec S12 24.69953 77. 74343 Aug 6 97 1.5 11 6.5 ... S. Long Rock S13 24.63067 77.691 Aug 9 97 2.5 15 9.0 ... Mid Long Rock SI 4 24.6259 77.69318 Aug 8 97 1.5 24 8.0 ... Sugar Rock Central Andros- S15 24.5448 77.68372 Aug 10 98 2.0 8 7.0 20.0 ± 11.0 all' 114 7.8 38 Autec2-South S16 24.484 77.699 Aug 1 1 98 1.0 10 7.0 24.0 ± 11.5 N. Bight S17 24.4395 77.69807 Aug 12 98 2.0 11 9.0 33.5 ±9.0 Big Wood S18 24.36703 ii.ems Aug 13 98 1.5 12 9.0 36.0 ± 11.0 Autec 3 S19 24.34315 77.67068 Aug 12 98 1.0 11 10.0 39.5 ± 15.5 Middle Bight S20 24.3069 77.65638 Aug 13 98 1.5 11 9.5 35.5 ± 10.5 Mangrove C. S21 24.29167 77.6462 Aug 13 98 1.5 15 8.5 25.0 ± 10.5 Mangrove S S22 24.25315 ii.m\i Aug 13 98 1.0 13 10.0 30.0 ± 13.5 Bights-all' 83 9.0 28.5 Congo Town S23 24.3013 11.6A1U Aug 15 98 1.0 15 9.0 30.5 ± 14.5 Long Bay Cay S24 24.09793 77.53703 Aug 16 98 1.0 12 10.5 35. 5± 11.5 North Rock S25 23.79092 77.42637 Aug 14 97 2.0 15 9.5 ... North Grassy S26 23. 77822 77.41902 Aug 13 97 2.0 15 9.0 ... Delta S27 23. 70967 77.38237 Aug 13 97 2.0 13 9.5 ... Pigeon South Andros- S28 23.6965 77.377 Aug 11 97 1.5 14 8.5 ... all' 84 9.3 33.0 Transects (#) = sum; all other columns = means ^Based on analysis of the first 10 transects/site. 95 Table IB. Fore reef site information for AGRRA stony coral and algal surveys off Andros Islanc , Bahamas {1997 sites are italicized). Deep Fore Reef Site Name Site Code Latitude (N°') Longitude (W“') Survey Date Depth (m) Transects (#) >25 cm Stony Corals (#/10mf % Live Stony Coral Cover (mean ± sd)' N. Joulters Dl 25.3132 78.0856 Aug 16 97 6.5 12 16.0 ... Nichols D2 25.1438 72.9875 Aug 17 97 9.5 13 13.0 ... Conch D3 25.1026 77.9639 Oct 17 97 5.5 9 8.5 ... Bucket D4 24.8793 77.8784 Oct 20 97 10.0 6 11.0 ... N.S laniard D5 24.8775 77.8791 Oct 18 97 7.5 13 6.5 ... S. Staniard 2 D7 24.50630 77.56437 Aug 20 98 9.0 9 11.0 37.0 ± 13.0 S.S laniard I D6 24.8438 77.9406 Oct 19 97 8.0 6 14.5 ... North Andros- all' 68 11.5 37.0 .S'. Love Hill D8 24. 7678 77.7987 Aug 6 97 8.5 16 9.0 ... Coffee D9 24. 7452 77. 7786 Aug 7 97 11.0 15 7.5 ... West Klein DIO 24.7450 77.7847 Aug 7 98 10.5 10 9.0 24.5 ±8.0 S. Aiilec DU 24. 7015 77. 7402 Aug 7 97 13.0 8 9.5 ... S. Long Rock Dll 24.6307 77.6910 Aug 9 97 8.0 12 11.5 ... Long Rock D13 24.6260 77.6910 Aug 8 98 8.7 10 11.0 22.5 ±8.0 Mid Long Rock DI4 24.37547 77.41587 Aug 9 97 9.5 11 10.0 ... Green Cay D15 24.5958 77.6933 Aug 9 98 11.0 14 3.0 11.0 ±4.0 Sugar Rock D16 24.5402 77.6821 Aug 10 98 11.5 13 6.5 18.0 ±7.0 Central Andros- all' \y Bristol Galley D17 24.5263 77.6892 Aug 10 98 11.5 12 7.0 17.0 ±5.0 Autec 2 D18 24.5064 77.6970 Aug 1 1 98 12.5 15 3.5 9.0 ±3.0 Autec 2-South D19 24.4833 77.6965 Aug 1 1 98 10.5 12 7.5 15.0±5.0 N. Bight D20 24.4393 77.6962 Aug 12 98 9.5 11 11.5 31.0± 12.5 Autec 3 D21 24.3432 77.6707 Aug 14 98 10.5 10 9.0 25.5 ± 10.5 Middle Bight D22 24.3097 77.6530 Aug 14 98 9.0 10 11.0 28.0 ±8.0 Mangrove N. D23 24.3011 77.6478 Aug 14 98 9.5 23 2.0 5.5 ±3.0 Bights-all' 93 7.4 18.5 Congo Town D24 24.3012 77.6479 Aug 15 98 10.5 10 12.0 21.0± 15.5 Long Bay Cay D25 24.0997 77.5334 Aug 16 98 10.5 10 11.0 19.0 ±9.0 Oasis D26 23.9476 77.3867 Aug 16 98 6.0 10 12.5 45.5±2L0 High Point Cay D27 23.4200 77.4600 Aug 17 98 9.0 9 13.5 35.0 ±7 .5 North Rock D28 23. 7965 77.4222 Aug 13 97 7.5 11 11.5 ... North Grassy D29 23. 7803 77.4168 Aug 13 97 9.5 9 10.0 ... South Grassy D30 23. 7285 77.3974 Aug 12 97 7.5 10 12.5 ... Delta D3I 23.7097 77.3810 Aug 13 97 8.5 12 7.5 ... Pigeon D32 23.6945 77.3742 Aug 11 97 9.0 12 8.0 ... Saddleback D33 23.6767 77.3703 Aug 11 97 8.5 10 12.5 ... South Andros- all' 103 11.2 30.0 Transects (#) = sum; all other columns = means ^Based on analysis of the first 10 transects/site. 96 Table 2A. Size and condition (mean ± standard deviation) of all stony corals (>25 cm diameter) in first 10 transects/site, by reef-crest sites off Andros {1997 sites are italicized). Reef Crest Site Year Stony Corals Partial-Colony Mortality (%) Stony Corals (%' Site Name Code (#) Diameter (cm) Recent^ Old Total Standing dead Bleached Diseased N. Jonllers SI 1997 177 114.5 ±5.5 6.5 ±0.5 21.5 ±2.0 28.0 ±2.0 5.5 1.5 8 Golding S2 1997 144 117.5 ±6.0 5.5 ±0.5 33.0 ±3.0 38.5 ±3.0 10.5 2 5 Morgan S3 1997 144 114.5 ±7.0 7.0 ±1.5 21.5 ±2.0 28.0 ±2.5 5.4 2 7 Coconut Point S4 1997 138 134.0 ±6.0 6.5 ±1.0 37.0 ±3.5 42.0 ±3.0 20.5 0 3 Mahore S5 1998 94 123.0 ±9.5 2.0 ±0.5 12.0 ±2.0 13.5 ±2.0 1 14 6.5 S. Staniard 2 S7 1998 64 135.5± 12.5 0.5 ±<0.5 24.0 ±4.0 24.5 ±4.0 8 0 15 S. Staniard ! S6 1997 80 191.0 ±12.5 4.5 ±1.0 43.5 ±4.5 47.0 ±4.5 24.5 11 0 North Andros- all' 841 133 ±27 4.5 ± 2.5 27.5 ±11.0 31.5 ±11.5 10.5 ±8.8 4.5 ± 5.7 6.5 ± 4.7 N. Love Hill S8 1998 85 101.5 ± 10.0 3.5 ± 1.5 24.0 ±2.5 27.0 ±3.0 0 3.5 1 5. Love Hill S9 1997 116 116.0 ±6.5 11.0±1.5 38.0 ±3.0 47.5 ±3.0 8.5 6 12 China Point SIO 1998 109 105.5 ±8.0 2.5 ±0.5 25.0 ±3.0 27.0 ±3.0 3.5 10 4.5 Red Rock Sll 1998 108 97.0 ±5.5 2.5 ±0.5 32.0 ±2.5 35.0 ±3.0 1 7.5 25 5. Autec S12 1997 71 128.5 ±9.0 9.5 ±1.5 41.5 ±4.5 50.0 ±4.0 14 1.5 25.5 S. Long Rock S13 1997 130 126.0 ±9.0 8.5 ±1.0 37.5 ±2.5 45.5 ±3.0 8.5 1 14 Mid Long Rock SI 4 1997 195 108.0 ±4.5 8.5 ±0.5 36.5 ±2.0 44.0 ±2.5 7. 3.5 9.5 Sugar Rock S15 1998 57 117.5 ±90 4.0± 1.0 27.5 ±4.5 31.5±4.0 9. 0 14 Central Andros all' 871 112.5± 11.5 6.0 ±3.5 33.0 ± 6.5 38.5 ± 9.5 6.5 ± 4.6 4.1 ±3.5 13.2 ±8.7 Autec 2-South S16 1998 69 96.0 ±6.5 1.0 ±0.5 39.0 ±4.5 39.5 ±4.5 16 4.5 10 N. Bight SI7 1998 99 117.5 ±6.0 2.0 ±0.5 31.5±3.5 33.0 ±3.5 13 1 8 Big Wood SIS 1998 109 89.0 ±5.5 4.5 ±0.5 23.0 ±3.0 27.0 ±3.0 9 7.5 2 Autec 3 S19 1998 no 85.0 ±5.0 3.5 ±0.5 25.0 ±3.0 28.5 ±3.0 7.5 8 3.5 Middle Bight S20 1998 103 88.0 ±6.0 5.0± 1.0 35.0±3.5 39.5 ±3,5 12.5 5 1 Mangrove C. S21 1998 124 84.0 ±5.0 3.5 ± 1.0 38.5 ±3.5 41.0±3.5 17.5 13.5 4 Mangrove S. S22 1998 130 73.0 ±4.5 2.5 ±0.5 25.0 ±3.0 27.5 ±3.0 9 3 6 Bights-all' 744 90.5 ± 14.0 3.0 ± 1.5 31.0 ±6.5 33.5 ± 6.0 12.0 ±3.7 6.1 ±4.1 4.9 ± 3.2 Congo Town S23 1998 138 93.5 ±4.5 3.0 ±0.5 30.0 ±3.0 33,0 ±3.0 13 8.5 3.5 Long Bay Cay S24 1998 124 88.0 ±5.0 15.0 ±2.5 28.0 ±3.0 41.0±3.0 3 80 1.5 North Rock S25 1997 146 119.0 ±7.0 7.5 ±1.0 52.0 ±3.0 55.5 ±2.5 18 0.5 7 North Grassy S26 1997 132 109.5 ±7.0 6.0 ±1.0 47.5 ±3.0 53.0 ±3.0 12 2.5 13 Delta S27 1997 123 119.0 ±6.0 3.0 ±0.5 40.0 ±3.0 43.0 ±3.0 4 1 4 Pigeon S28 1997 117 131.5 ±8.5 7.5 ±1.0 41.0 ±3.0 48.5 ±3.0 7 0 31 South .Andros 780 1 10.0 ± 16.5 7.0 ± 4.5 40.0 ± 9.5 46.0 ± 9.0 9.5 ± 5.8 15.4±31.8 10.0± 11.0 all' 'stony corals (#) = sum; all other columns = means ± standard deviation. ^1997 values are inflated relative to 1998 due to ehanges in assessment methodology. Table 2B. Size and condition (mean ± standard deviation) of all stony corals (>25 cm diameter), in first 10 transects/site, by fore-reef sites off Andros {1997 sites are italicized). 97 Fore-reef Site Year Stony corals Partial-colony mortality (%) Stony corals (%) Site name Code (#) Diameter (cm) Recent^ Old Total Standing dead Bleached Diseased N. Joulters DI 1997 166 65.0 ±2.5 11.0±1.5 21.0±2.0 32.0 ±2.0 2.5 0.5 1 Nichols D2 1997 168 72.0 ±4.0 11.0±1.5 33.0 ±2.5 43.0 ±2.5 7 3 2 Conch D3 1997 77 57.5 ±3.5 6.0 ±1.5 19.0±3.0 25.0 ±3.0 1.5 2.5 1.5 Bucket D4 1997 58 70.5 ±7.0 2.0 ±0.5 30.0 ±3.5 32.0 ±3.5 0 14 1.5 N. Staniard D5 1997 77 60.5 ±5.0 7.0 ±1.5 39.5 ±4.0 46.0 ±4.0 6.5 4 1.5 S. Staniard 2 D6 1998 99 54.0 ±4.0 4.0 ± 1.5 16.5±2.0 20.5 ±2.5 2 8 1 S. Staniard 1 D7 1997 87 59.5 ±3.5 8.0 ±1.5 33.5 ±3.5 40.5 ±3.5 5.5 10.5 1 North Andros- all' 732 62.5 ± 6.5 7.0 ± 3.5 27.5 ±8.5 34.0 ± 9.5 3.5 ± 2.7 6.0 ± 4.9 2.2 ± 2.1 5 .Love Hill D8 1997 98 69.0 ±3.5 12.0 ±2.0 37.5 ±3.5 48.5 ±3.5 2 1 12 Coffee D9 1997 113 55.5 ±3.0 16.5 ±2.0 31.0±3.0 46.5 ±3.0 4.5 2.5 9 West Klein DIO 1998 89 47.5 ±2.0 24.5 ±3.5 16.0±3.0 39.0 ±3.5 3.5 13.5 42.5 S. Autec Dll 1997 74 66.5 ±4.0 19.0 ±2.5 30.0 ±3.0 48.0 ±3.5 2.5 4 2.5 S. sLong Rock D12 1997 135 59.5 ±2.5 14.0 ±1.5 25.0 ±2.0 39.0 ±2.5 0 5 8 Long Rock D13 1998 no 58.5 ±3.0 39.0 ±4.0 27.0±3.5 58.5 ±3.5 7.5 11 27.5 Mid Long Rock D14 1997 110 55.0 ±3.0 13.5 ±1.5 22.5 ±2.0 36.0 ±2.5 1 5.5 7.5 Green Cay D15 1998 37 41.0±3.5 15.5 ±5.0 13.0±3.5 28.0 ±5.5 2.5 2.5 13.5 Sugar Rock D16 1998 78 34.0 ± 1.0 8.5 ±2.5 20.5 ±3.0 28.5 ±3.5 1.5 5 16.5 Central Andros-all' 844 54.0 ± 11.5 18.0 ±9.0 24.5 ±7.5 41.5 ± 10.0 2.7 ± 2.2 5.5 ±4.1 15.4 ± 12.4 Bristol Galley D17 1998 74 38.5 ±2.0 15.0 ±3.5 24.0 ±4.0 37.5 ±4.5 7 4 23 Autec 2 D18 1998 46 39.0 ±3.0 8.5 ±4.0 22.5 ±5.5 29.0 ±5.5 11 11 6.5 Autec 2-South D19 1998 81 46.5 ±3.5 n.o±3.o 29.0 ±4.0 38.0 ±4.0 8.5 0 11 N. Bight D20 1998 115 54.5 ±3.5 12.5 ±2.5 20.0 ±2.5 30.5 ±3.0 5 2.5 15.5 Autec 3 D21 1998 89 39.5 ±2.0 14.5 ±3.0 20.5 ±3.0 34.0 ±3.5 2 8 21.5 Middle Bight D22 1998 106 5L5±3.5 16.0 ±3.0 22.5 ±3.0 37.5 ±3.5 4. 2 22.5 Mangrove N. D23 1998 47 34.0± 1.5 7.0 ±2.0 17.5 ±3.5 24.5 ±3.5 0 6.5 6.5 Bights-all' 558 43.5 ± 7.5 12.0 ±3.5 22.5 ±3.5 33.0 ± 5.0 5.4 ± 3.8 4.8 ± 3.8 15.2 ±7.3 Congo Town D24 1998 122 53.5 ±2.5 42.5 ±3.5 25.5 ±3.0 62.0 ±3.0 2.5 2.5 32 Long Bay Cay D25 1998 no 43.0 ±2.0 20.5 ±3.5 23.0 ±3.0 40.0 ±3.5 5.5 4.5 25.5 Oasis D26 1998 127 72.5 ±5.0 2.0 ±0.5 26.0 ±2.5 28.0 ±2.5 2.5 1.5 3 High Point Cay D27 1998 114 45.5 ±2.0 5.0± 1.5 20.5 ±2.5 25.5 ±2.5 2 2.5 8 North Rock D28 1997 124 63.5 ±3.5 9.0 ±1.5 29.5 ±2.5 38.0 ±2.5 1.5 0 1.5 North Grassy D29 1997 92 54.0 ±4.0 8.5 ±1.5 33.5 ±3.0 42.0 ±3.0 1 2 2 South Grassy D30 1997 125 69.0 ±3.5 10.0 ±1.0 39.0 ±2.5 49.0 ±2.5 5 1 4 Delta D31 1997 90 53.0 ±2.5 11.0 ±2.5 31.5±3.0 42.0 ±3.5 3.5 1 3.5 Pigeon D32 1997 97 68.5 ±4.5 13.5 ±2.5 45.0 ±4.0 56.0 ±3.5 12.5 1 6 Saddleback D33 1997 123 63.5 ±3.0 11.0±1.5 29.5 ±2.5 40.5 ±2.5 2.5 4 12 South Andros- all' 1124 58.5 ± 10.0 13.5 ± 11.5 30.5 ±7.5 42.5 ± 11.0 3.8 ± 3.3 2± 1.4 9.8 ± 10.6 Stony corals (#) = sum; all other columns = means ± standard deviation. ^1997 values are inflated relative to 1998 due to changes in assessment methodology. 98 Table 3 A. Algal characteristics, and density of stony coral recruits and of Diadema antillarum (mean ± standard deviation), by reef-crest sites off Andros {1997 sites are italicized). Reef Crest Site Name Site Code Year Quadrats {#) Relative Abundance (%) Macroalgae Turf algae Crustose coralline algae Macroalgal Height^ Index^ Recruits Diadema (#/.0625 mQ (#/100mQ N. Joulters SI 1997 53 6.5 ±10.0 43.0 ±11.0 50.5 ±6.0 ... ... 0.2 ±0.4 0 Golding S2 1997 40 7.5 ± 12.0 39.5 ±13.5 53.0 ±11.5 ... ... 0.0 ±0.2 0 Morgan S3 1997 44 8.0 ±8.5 51.5 ±16.5 40.5 ±18.0 ... ... 0.3 ± 0.5 0 Coconut Point S4 1997 41 8.5 ± 15.0 44.0 ±16.5 47.5 ± 14.5 ... — 0.1 ±0.3 0 Mahore S5 1998 45 12.0 ± 16.0 49.5 ± 18.5 38.5 ± 17.0 3 82 0.1 ±0.4 1 S. Staniard 2 S7 1998 34 13.0 ±7.5 64.5 ±22.0 22.5 ± 19.5 2 41 0.1 ±0,3 0 S. Staniard / S6 1997 36 3.5 ±6.0 49.5 ±16.0 47.0 ±15.5 ... ... 0.2 ± 0.3 0 North 293 8.0 ± 3.0 49 ± 8.0 43.0 ± 10.5 2.5 ± 0.5 61.5 ± 0.2 ±0.1 0.1 ± 0.4 Andros-all' 29.0 N. Love Hill S8 1998 40 17.0 ±23.0 39.5 ±25.0 43.5 ±27.0 2.5 79 0.4 ±0.8 0 5. Love Hill S9 1997 60 18.5 ±15.5 41.0 ±27.5 40.5 ±29.5 ... — 0.3 ±0.7 0 China Point SIO 1998 44 10.5 ±9.5 37.0 ±25.5 52.5 ±26.5 2 28 0.1 ±0.5 0 Red Rock Sll 1998 18 13.5±21.5 57.0 ±26.0 29.5 ±27.0 1 15 0.1 ±0,2 0 S. Auiec SI2 1997 42 26.0 ±21.0 38.0 ±22.5 35.5 ±24.0 ... ... 0.3 ±0.5 0.9 S. Long Rock SI 3 1997 46 14.5 ±14.0 40.5 ± 14.5 45.0 ±19.0 ... ... 0.2 ±0.4 0 Mid Long Rock SN 1997 92 9.5 ± 10.5 50.5 ±24.5 39.5 ±26.0 ... ... 0.4 ±0.6 0 Sugar Rock S15 1998 37 8.0 ±8.5 50.0 ±20.5 42.0 ±22.0 1.5 23 0.2 ±0.8 0 Central 379 14.0 ±5.5 44.0 ± 7.0 42 ± 8.0 1.5 ±1.0 36.5 ± 0.3 ±0.1 0.1 ±0.3 Andros-all' 29.0 Autec 2-South S16 1998 46 10.5± 11.0 44.5 ±23.0 44.5 ±25.0 1 24 0,0 ±0.2 0 N. Bight S17 1998 55 13.5 ±20.0 45.0 ±21.5 39.0± 18.0 1.5 26 0,1 ±0.6 0 Big Wood S18 1998 54 26.0 ±20.5 37.0 ±23.5 36.5 ±26.0 2.5 99 0.1 ±0.4 4 Autec 3 S19 1998 54 9.0 ± 12.0 46.5 ±25.0 44.5 ±28.5 1.5 22 0.1 ±0.7 0 Middle Bight S20 1998 54 13.5 ±20.0 40.0 ±21.0 46.0 ±22.5 1.5 39 0.1 ±0.3 2,5 Mangrove C. S21 1998 65 10.5 ± 13.5 40.5 ±21.0 48.5 ±22.0 1.5 47 0.2 ±0.5 0 Mangrove S S22 1998 64 20.5 ± 17.5 42.0 ± 18.0 37.5 ± 18.0 2.5 88 0.2 ±0.5 0 Bights-all' 392 15.0 ±6.0 42.0 ± 3.5 42.5 ± 4.5 1.5 ±0.5 49.0 ± 0.1 ±0.1 1 ± 1.5 31.5 Congo Town S23 1998 69 1L5± 15.5 13.5 ± 19.0 44.0 ± 19.0 1.5 26 0.1 ±0.4 2.5 Long Bay Cay S24 1998 54 1L0± 14.5 14.0 ± 19.0 41.0± 19.5 2 43 0.1 ±0.3 0 North Rock S25 1997 50 14.5 ±19.0 29.5 ±21.0 56.0 ±16.0 ... ... 0.1 ±0.4 0 North Grassy S26 1997 51 10.5 ±11.5 43.0 ±19.5 47.0 ±17.5 ... ... 0.4 ±0.7 0 Delta S27 1997 44 11.5 ±12.0 30.5 ±15.0 58.0 ±16.5 ... ... 0.3 ±0.5 0 Pigeon S28 1997 44 18.0 ±15.0 40.5 ±15.0 41.5 ±18.0 ... ... 0.2 ±0.4 0 South 312 12.5 ±3.0 28.5 ± 12.5 48.0 ± 7.5 1.5 ±0.5 34.5 ± 0.2 ±0.1 0.5 ± 1.0 Andros-all' 12.0 'Quadrats (#) = sum; all other columns = means ± standard deviation. ^Heights measured as maximum canopy heights in 1997 are not included. ^Macroalgal index = relative abundance of macroalgae x macroalgal height. Table 3B. Algal characteristics, and density of stony coral recruits and of Diadema antillarum (mean ± standard deviation), by fore-reef sites off Andros {J997 sites are italicized). 99 Fore Reef Site Year Quadrat Relative Abundance (%) Macroalgal Recruits Diadema Site Name Code m Macroalgae Turf algae Crustose Height Index^ (#/.0625 (#/100 coralline 2 mQ mQ algae N. Joulters D1 1997 35 49.0 ± 18.0 10.0± 13.5 41.0± 18.5 ... ... 0.3 ±0.6 0 Nichols D2 1997 42 35.5 ± 15.0 17.0± 13.0 47.5 ± 14.0 ... — 0.5 ±0.6 0 Conch D3 1997 30 60.0 ±32.0 26.5 ±26.5 14.0 ± 14.0 ... ... 0.3 ±0.6 0 Bucket D4 1997 24 25.5 ±8.5 45.0 ±22.0 29.5 ± 18.5 ... ... 0.7 ±0.8 0 N.Staniard D5 1997 21 33.5 ± 18.5 44.0 ±26.0 22.5 ±8.0 ... ... 0.5 ±0.6 0 S. Staniard 2 D6 1998 45 35.0 ± 14.0 27.5 ± 18.5 37.5 ± 19.0 3.5 162 0.4± 1.1 1 5. Staniard 1 D7 /997 24 36 ± 10.0 46.0 ± 18.0 22.5 ± 18.0 ... ... 0.8 ± 1.0 0 North Andros- all' 22! 39.0 ± 11.5 31.0 ± 14.0 30.5 ± 12.0 3.5 162 0.5 ± 0.2 0.2 ± 0.4 S .Love Hill D8 1997 37 61.0± 19.5 25.0 ± 14.5 14.5 ± 12.0 ... ... 0.9 ±1.2 0 Coffee D9 1997 51 56.5 ± 17.0 13.5 ± 14.5 30.0 ± 14.0 ... ... 0.7 ±0.9 0 West Klein DIO 1998 45 48.0 ± 19.5 28.0 ± 18.0 24.0 ± 11.5 3.5 177 0.5 ±0.9 0 S .Autec Dll 1997 33 74.0 ± 17.5 6.5 ±9.5 19.0 ± 16.0 ... ... 0.7 ± 1.0 0 S. Long Rock D12 1997 38 62.5 ±20.5 10.5± 11.5 27.0 ± 16.0 ... ... 0.9 ±0.9 0 Long Rock D13 1998 46 48.0 ±20.0 30.0 ± 19.5 22.0 ± 13.0 3.0 145 0.1 ±0.4 0 Mid Long Rock D14 1997 33 59.0 ± 15.0 9.5 ± 12.0 31. 5± 10.5 ... ... 0.7 ±0.8 0 Green Cay D15 1998 62 54.5 ±20.0 15.5 ± 19.0 30.0 ± 14.0 3 212 0.7 ±0.8 0 Sugar Rock D16 1998 56 41.5 ± 12.5 21.0± 16.5 37.5 ± 15.0 3.5 215 0.9±1.0 0 Central Andros all' 401 56.0 ± 9.5 18.0 ±9.0 26.0 ± 7.0 3.2 ± 0.3 187 ±33 0.7 ± 0.2 0 Bristol Galley D17 1998 48 45.5 ±21.0 18.0 ±20.5 36.5 ± 15.0 4.0 238 0.8± 1.1 0 Autec 2 D18 1998 62 51.5 ± 13.0 11.0 ±9.0 37.5 ± 10.5 3.5 281 0.7±1.1 2 Autec 2-South D19 1998 44 40.0 ± 15.5 24.0 ±24.0 36.0 ± 18.5 3.0 178 0.4 ±0.7 0 N. Bight D20 1998 47 45.0 ± 15.5 21.0 ±22.5 33.5 ± 16.0 3.0 202 0.8±1.2 0 Autec 3 D21 1998 49 44.0 ± 16.0 14.0± 17.5 42.0 ± 18.0 3.0 185 0.6 ±0.7 0 Middle Bight D22 1998 48 42.0 ± 13.0 14.0± 15.5 44.0 ± 14.0 3.7 255 0.6 ± 1.0 0 Mangrove N. D23 1998 70 50.5 ± 19.0 15. 5± 17.0 34.0 ±22.0 3.0 238 0.2 ±0.6 0 Bights-all' 368 45.5 ± 4.0 17.0 ±5.0 37.5 ± 4.0 3.3 ± 0.4 226 ± 38 0.6 ± 0.2 0.3 ± 0.7 Congo Town D24 1998 48 50.0 ±24.0 13.5 ± 19.0 36.5 ±23.0 4.0 218 0.3 ±0.7 0 Long Bay Cay D25 1998 50 50.0 ± 16.5 14.0 ±1 9.0 36.0 ± 17.5 3.5 223 0.4 ±0.7 4 Oasis D26 1998 50 31. 5± 19.5 29.0 ±25.5 39.5 ± 17.5 3.0 140 0.6 ± 1.2 0 High Point Cay D27 1998 35 41.5± 16.0 16.0±21.5 42.5 ± 19.5 3.5 190 0.9 ± 1.2 2 North Rock D28 1997 33 49.5 ± 12.5 3.5 ±6.0 47.5 ± 13.0 ... ... 1.1 ± 1.5 0 North Grassy D29 1997 26 43.0 ±25.0 10.0 ± 16.0 47.0 ±24.0 ... ... 0.9± 1.8 0 South Grassy D30 1997 31 49.0 ± 15.0 4.0 ±7.0 46.5 ± 12.0 ... ... 0.6 ±0.8 0 Delta D31 1997 43 41.0± 15.0 4.0 ±6.5 55.0 ± 16.0 ... ... 1.1 ± 1.3 0 Pigeon D32 1997 39 48.5 ± 15.5 9.0 ± 11.0 42.5 ± 12.5 ... ... 1.7± 1.7 0 Saddleback D33 1997 37 48.0 ± 16.0 5.5 ±9.0 46.5 ± 15.0 — ... 0.9 ± 1.1 0 South Andros- all' 392 45.5 ± 6.5 11.0 ±8.0 38.5 ± 3.0 3.3 ± 0.4 193 ±38 0.9 ± 0.4 0.6 ± 1.4 'Quadrats (#) = sum; all other columns = means ± standard deviation. ^Heights measured as maximum canopy heights in 1997 are not included. ^Macroalgal index = relative abundance of macroalgae x macroalgal height. 100 Figure 1. Location of Andros Island, the Bahamas (inset) and the 1997 and 1998 AGRRA survey sites in four areas (North, Central, Bights, South) off eastern Andros. See Tables lA, IB for site codes. ASSESSMENT OF ANDROS ISLAND REEF SYSTEM, BAHAMAS (PART 2: FISHES) BY PHILIP A. KRAMER,* KENNETH W. MARKS, ^ and TIMOTHY L. TURNBULL^ ABSTRACT Coral reef fish assemblages were surveyed at 48 reef-crest and fore-reef habitats along approximately 200 km of reefs on the eastern side of Andros Island in August of 1997 and 1998. A total of 164 species were recorded in roving diver surveys, averaging 55 species per site. Select species density averaged 37.4 individuals/ lOOm^ in belt transects and was significantly more abundant in reef crests than fore reefs. The select fish assemblages were dominated by scarids, haemulids, and acanthurids, while serranids were ubiquitous but present in low densities (<0.5/1 OOm^). Small differences in the community structure of four geographic areas (north, central, bights, south) are indicative of well-mixed populations. Species richness and abundance were comparatively low, particularly in fore-reef habitats, although mean size and biomass were relatively high. The Andros reef fish assemblages may be naturally limited by low recruitment, lack of nursery habitat, or possibly by high levels of predation. The entire reef system may be at high risk to even modest increases in fishing. INTRODUCTION Intact fish assemblages are integral to the functioning of coral reefs and patterns in their diversity, abundance, and size can be used to understand underlying ecological processes such as recruitment, predation, and herbivory. These patterns vary at different spatial and temporal scales and are likely to be influenced by habitat variables such as topographic complexity (e.g., Connell and Kingsford, 1998; Nunez-Lara and Arias- Gonzalez, 1998), live coral cover (e.g.. Bell and Galzin, 1984), depth (e.g., Lewis and Wainwright, 1985), wave energy (McGehee, 1994; Mejia and Garzon-Ferreira, 2000), * Marine Geology and Geophysics, Rosensteil School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: pkramer@rsmas.miami.edu 2 Reef Environmental Education Foundation, 98300 Overseas Highway, Key Largo, FL, 33037. Email: kema@adelphia.net ^ Andros Conservancy and Trust, P.O. Box 21667, Ft. Lauderdale, FL 33335. Email: ttumbullspyglass@att.net 102 and cross shelf position (Lindeman et al., 1998) as well as species interactions (e.g., predation, competition), suitability of substratum (e.g., algal cover, Lawson et al., 1999), and larval availability and recruitment patterns (e.g., Cowen et al., 2000). The application of comprehensive habitat-based sampling programs (e.g., Ault et al., 2001) has proven effective in gaining insight into complex patterns of fish assemblages. The spatial variation in the abundance and distribution of fish communities off Andros Island, Bahamas, one of the most extensive reef systems in the Western Atlantic, is poorly known. The reef tract parallels the eastern side of the island extending 217 km from Joulters Cays in the north to Saddleback Cays in the south. Shallow reef crests and outer slope reefs are the principal reef types although lagoonal patch reefs are also present. The presence of extensive, topographically complex reef crests and fore reefs (Kramer et al., this volume) would tend to suggest that suitable habitat is sufficiently available to support abundant and diverse fish populations off Andros Island. Historically, fishing has been mostly local, artisanal-level fishing concentrated in the central and southern regions of Andros, with relatively little commercial activity. Harvesting has probably altered fish communities less significantly than reported for other areas in the Caribbean (e.g., Roberts, 1995). However, in the past decade commercial fishing has increased, with larger operations particularly active south of the bights in Mangrove Cay and South Andros. Significant fishing effort is concentrated in December and January around several large Nassau grouper (Epinephelus striatus) spawning aggregations located in central and southern Andros. Little data are available on fishing pressures, since statistics on landing sites and total fishing effort are not recorded (Bahamas Reef Environmental Education Foundation and Macalister Elliot and Partners, 1998, unpublished report). In this paper, we present the results from a large-scale (>100 km) habitat-based assessment of the fishes off Andros Island reefs using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) methodology and examine how elements of species richness, density, and biomass vary spatially and between habitat types. Information on the status of fish communities can provide an essential baseline critical for management and conservation efforts of fishes, particularly targeted species. Spatial trends and the condition of principal reef-building corals and algal populations are presented in Kramer et al. (this volume). METHODS Andros Island is located in the central Bahamas where the Great Bahamas Bank meets the Tongue of the Ocean (Fig. 1). An extensive yet discontinuous fringing bank barrier reef parallels the eastern side of the island. Reef crests are dominated by colonies of Acropora palmata (live and standing dead) and display varying degrees of development controlled, in part, by wave energy, reef aspect, and the presence of freshwater creeks. Fore reefs range from hard-bottom assemblages dominated by gorgonians to reefs composed of dense coral growth dominated by the three morphotypes of the Montastraea annularis species complex, which reach heights of 2-3 m off the 103 bottom. Structurally developed fore reefs are often associated with well-developed reef crests and occur at “intermediate” depths of 7-12 m. The Andros reef tract (Fig. 1) was divided into four geographic areas: north (N); central (C); bights (B); and south (S). Fish surveys were conducted at the same locations as the benthic surveys (Kramer et al., this volume) and are representative of the better- developed reefs within two stratified habitat types (shallow reef crest and intermediate- depth fore reef). In 1997, 17 sites (8 at 1-3 m depth, 9 at 8-12 m) located mainly in the northern and southern areas were surveyed. In 1998, the 31 surveyed sites (15 shallow and 16 deep) were located in all four areas (Tables lA, IB). At each site, a combination of belt transects and roving diver surveys were used to assess the fish community structure using the AGRRA methodology at the same time that benthic characteristics [including live stony coral cover, density of “large” (>25 cm diameter) stony corals, relative algal cover] were evaluated. In 1997, the AGRRA Version 1.0 fish protocol was employed, except that only 2- 5 belt transects (each 50 x 2 m) were made at each site because our underwater time was limited. All species and all sizes of haemulids (grunts), scarids (parrotfishes), and serranids (groupers) present in the belt transects were counted during these surveys. In 1998, when AGRRA Version 2.0 (see Appendix One, this volume) was used, 10 belt transects, each 30 x 2 m, were deployed at each site except at four locations (S5, SI 1, S20, DIO having 6, 4, 5, and 9 transects, respectively). In 1998, counts of serranids were restricted to species of Epinephelus and Mycteroperca, while scarids and haemulids less than 5 cm in length were not tallied, but the number of fish species quantified in the belt transects was expanded to include bar jack {Caranx ruber), yellowtail damsel {Microspathodon chrysurus), barracuda {Sphyraena barracuda), hogfish (Lachnolaimus maximus), Spanish hogfish (Bodianus rufus) and all balistids. In 1998, swimming speeds were 6-8 minutes per 30 m transect, while in 1997 they were 8-10 minutes per 50 m transect. Two of the authors (Marks, Turnbull) conducted the fish transects both years, and all roving diver species richness counts were done by the same surveyor (Marks). Fish identification was based on Humann (1994). Statistical analyses were performed with the program Statistica (Version 5.1). Transect averages of fish density were calculated for each site based on the number of transects deployed and represented as the mean number per transect adjusted to a common unit area of 100 m^. Parameters were analyzed by students t-test and by 1- and 2-way Analysis of Variance (ANOVA). The four geographic areas (N, C, B, S) were analyzed by ANOVA with sites hierarchically nested under areas as random factors. Density and biomass data were checked to ensure that variances met assumptions of homoscedasticity. Regression analysis was used to examine relationships between fish data and benthic habitat variables (coral cover, coral size, coral frequency, depth). RESULTS A total of 164 species, plus several other unidentified species of silversides, herrings, and anchovies, were documented during the roving diver surveys for the entire Andros reef tract (http://www.reef org). For the reef-crest sites, a total of 126 species were 104 observed during ~30 hours of roving bottom time (Table 1 A). In the fore reefs, 144 speeies were observed during ~33 hours of bottom time (Table IB). The average number of species encountered per site was 54 in the reef crests (average roving bottom time =78 minutes, n=23 sites) and 56 in the fore reefs (average roving bottom time=79 minutes, n=25 sites). Total species richness within an area (N, C, B, S) ranged from 113-128; approximately 60% of all recorded species were seen in each geographic area. Differences were found in the presence or absence of rare species with low sighting frequencies. For example, several species only seen in northern Andros included black durgon (Melichthys niger), trunkfish (Lactophyrys trigonus), and spotted trunkfish (L. bicaudalis) whereas the diamond blenny (Malacoctenus boehlkei) and porkfish (Anisotrematus virginicus) were seen in the bights and southern areas but not in the central or northern areas. A total of 50 fish species were counted (out of a possible 72 AGRRA species) within the belt transects. Forty-four transect species were counted in the reef crests compared to 47 in the fore reefs. Ten AGRRA species not recorded in transects were seen during roving diver surveys at low (<30%) sighting frequencies. In 1998, an average of 21 .5 fish species was recorded within transects at each site (n=31 sites). Reef crests had slightly more species (mean=22.5, n=15 sites) on average than fore reefs (mean=20.5, n=16 sites). Species similarity between the four areas of Andros was high with over 75% of the AGRRA-listed transect species counted within at least three of the four areas. The total number of adult AGRRA fishes in the belt transects was 8,800 with 6,281 individuals counted in reef-crest habitats and 2,519 in fore-reef habitats (all sites and both years combined). Reef-crest communities were dominated by haemulids (37%), scarids (25%), acanthurids (20%), and lutjanids (14%), while fore reefs contained a higher proportion of scarids (46%) and substantially fewer haemulids (7%) (Fig. 2). Figure 2. Relative abundance of AGRRA fishes in reef crests and fore reefs pooled for both years off Andros, Bahamas. Mean fish density for surveyed transect species was 37.4 individuals/ 100 m (both habitats and both years combined). Mean family density and biomass for the surveyed 105 species are shown in Figure 3. Higher fish densities were recorded in the 1998 surveys (overall niean=47.7/ lOOm^) than in 1997 (overall mean=l 9.2/1 OOm^), Statistically significant differences were detected for both herbivorous fishes (scarids >5 cm, acanthurids, M. chrysurus) (t- test, df=46, t=-3.7, p<0.001) and for carnivores (hameulids >5 cm, lutjanids, select serranids) (t- test, df=46, t=-3.44, p<0.01). When the data for all sites and both years are combined, large, free-ranging herbivores (scarids >5 cm, acanthurids) had relatively high densities (19.5/100 m^), more than three times that of major piscivores (select serranids, luhanids) (4.6/100 m ) (Tables 2A,B). For both survey years (all densities in numbers/ lOOm^), the reef crests contained significantly more fish (mean=58.1, sd=45.3, n=23 sites) than were found in the fore reefs (mean=18.3, sd=7.8, n=25 sites) (t-test, df=46, t=-4.3, p<0.0001). Moreover, the highest recorded fish densities (>100/100m^) consistently occurred in reef crests (e.g., SI 1, SI 8, S23) while the lowest densities (<8/100m^) were in fore reefs (e.g., D6, D12, D16). Total AGRRA fish density patterns are strongly weighted by scarid, acanthurid, and haemulid densities, all of which were significantly higher in reef-crest habitats (scarids >5 cm, p<.001; acanthurids, p<0.00001; haemulids >5 cm, p<0.001). Families with fewer sightings, including the Chaetedontidae Serranidae, and Balistidae, all had higher densities in the fore reefs (Fig. 3). Figure 3. (A) Density (mean no. fish/100 ± standard deviation) and (B) biomass (mean g/100 m^ ± standard deviation) for AGRRA fishes, pooled for both habitats (reef crests and fore reefs) and both years (1997 and 1998) off Andros, Bahamas. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Sphyraena barracuda. 106 A comparison between the relative abundance of species observed in transects versus their abundance (calculated as the sum of density x sighting frequency) in the roving diver surveys is shown in Figure 4 for the 15 most commonly observed AGRRA species for both years combined. Several species of haemulids (e.g., French grunt, Haemulon flavolineatum) and lutjanids (e.g., mahogany snapper, Lutjanus. mahogoni) had clumped distributions with high concentrations only in several shallow sites. Scarids were well represented with a total of 10 species seen in belt transects in all surveys combined and a higher species richness in fore-reef sites. The two most abundant species were stripped parrotfish (Scarus croicensis) and stoplight parrotfish (Sparisoma vihde) (Fig. 5A). The sizes of most adult parrotfishes were considered “average” for their species with approximately 20% in the 31-40 cm class range (Fig. 5A). All three species of acanthurids were present off Andros (average site 'y ^ density=7.9/100 m , n=48 sites), with nearly twice as many encountered in reef crests (n=l,258 in 23 sites) as in fore reefs (n=652 in 25 sites). Blue tangs (Acanthurus coenileus) made up 61% of the acanthurids seen in the belt transects followed by ocean surgeons (A. bahianus) {25%), and doctorfish (A. chirurgus) (14%). Most acanthurids were in the 1 1-20 cm size class. Another potentially important herbivore, the Bermuda chub (Kyphosus sectatrix) was rare off Andros (sighting frequency of 3 1 % in the roving diver surveys) and usually occurred as isolated individuals. — (U 25 French grunt {Haemulon flavolineatum) Striped parrotfish {Scarus croicensis) Blue tang \ • • {Acanthurus coeruleus) Mahogany snapper {Lutjanus mahogoni) • • • • Stoplight parrotfish •• • • {Sparisoma vihde) • • • • • 0 50 too 150 200 250 300 350 Sighting frequency x density (roving diver) Figure 4. Relationship between relative abundance in belt transects and roving diver surveys (^density x sighting frequency in 1997 + 1998) for the 15 most commonly observed AGRRA fish species off Andros, Bahamas. 107 Figure 5. Relative species abundance and size frequency distribution in cm of (A) scarids >5 cm and (B) serranids in reef crests and fore reefs off Andros, Bahamas. 108 The yellowtail damselfish (M chrysurus) was observed at 96% of reef-crest sites (Table 3A, roving diver sighting frequency) at belt transect densities of 1.9/100 m^, but at only 26% of fore reef sites at densities of 0.1/100 m^. Other territorial damselfish including (in order of sighting frequency in the roving diver surveys) bicolor (Stegastes partitiis), threespot (S. planifrom) , dusky (S.fucus), longfm (S. diencaeus), and cocoa (S. variabilis) were similarly most abundant in reef-crest habitats. The roving diver sighting frequency for serranids was high on Andros (e.g., they were observed at nearly all sites). Within belt transects, serranids composed 0.5% of the AGRRA fishes counted in reef crests and 3% in fore-reef sites (Fig. 2). The most frequently encountered serranid species in the belt transects, in order of abundance, were tiger grouper {Mycteroperca tigris), coney {Epinephelus fidviis), Nassau grouper {E. striatus), graysby {E. cruentatus), and red hind {E. guttatus) (Fig. 5B). Yellowmouth grouper (M interstitialis) and yellowfm grouper (M venenosa) were present in fore-reef habitats only, and other serranids, including black grouper (M bonaci) and rock hind {E. adscensionis), were only seen during roving diver surveys. The density of serranids within transects was quite low, averaging only 0.39 individuals/100 m^ for all years and sites combined. Significant differences in serranid density were detected between the reef crests and fore reefs (t-test, df=46, t=-2.4 p<0.05) but not among the four areas (N, C, B, S) (1-way ANOVA, df=3, MS=0.38, F=2.6, p=0.06). The sizes of adult groupers counted in transects were large, with nearly 30% of all groupers being greater than 40 cm in length (Fig. 5B). Eleven species of grunts were observed (roving diver surveys), with greater mean densities in belt transects at reef crests (22.8/100 m ) than in fore reefs (1.7/100 m ). Ten species of snapper were observed during roving diver surveys; schoolmasters {Lutjams apodus) and yellowtail snappers {Ocyurus chrysurus) were seen most frequently (Table 3A, B). Snapper densities in belt transects averaged 7.1/100 m^ on reef crests and 1.5/100 m on fore reefs. Mean snapper density for the shallow sites was significantly (p<0.05) higher in the bights (14.8/100 m^) than in the other geographic areas. Within each of the four areas, variation in total fish density was higher at the within- and between-site scale, particularly for reef-crest habitats, than among the geographic areas. For example. Figure 6 shows the variability in transect abundance at three spatial scales for scarids and acanthurids. Most of the variation (~80%) occurred at the within-site scale, with the remainder at the between-site scale. No significant differences were detected among the four areas of Andros (N, C, B, S) for either total species density (1-way ANOVA, df=3, MS=826, F=0.6, p=0.63) or total fish biomass (1- way ANOVA, df=3, MS=390259, F=1.4, p=0.25). Only grouper biomass was statistically different among the four areas (1-way Anova, df=3, MS=373576, F=4.2, p<0.05) mainly because few were seen on northern reef crests. A significant positive relationship was found between chaetodontid density and live stony coral cover in fore reefs (p<0.01), but not in reef crests. Analysis of herbivore- macroalgal index relationships within each of the habitat types was significant (p<0.05) only among fore reefs and not reef crests. However, when the two habitats were combined a significant inverse relationship (p<0.001) was found between the macroalgal index (a proxy for macroalgal biomass) and herbivore biomass in 1998 (Fig. 7). The best fit line shown in the figure is a log-fit with reef-crest and fore-reef sites separately circled. A significant negative relationship also existed for depth and herbivore biomass 109 A Scaridae 80 ■ 10-M (n = 170) □ 3-M (n = 153) o o O o o 1 (N (N 'O so ^ so f-H CN CN B Acanthuridae # fish per transect (100 m^) # fish per transect (100 m ; C Scaridae D Acanthuridae 80 ■i 60 Q. E o u 40 20 * P<.05 >i<* P<.0 *** p<.001 C Between areas Between sites Within sites 80 - * P<.05 - ** P<.01 60 - P<,001 40 - 20 - Between areas Between sites Within sites Figure 6. Frequency distribution of fish density recorded within 100 transects for (A) scarids >5 cm and (B) acanthurids for both reef-crest and fore-reef habitats. Breakdown of the components of variation in fish transect densities at three spatial scales (within a site, between sites, and between areas) for (C) scarids >5 cm and (D) acanthurids in both reef-crest and fore-reef habitats. *** = level of significance is indicated by number of stars. 110 Figure 7. Regression of mean herbivore biomass (acanthurids, scarids >5 cm, and Microspathodon chrysurus) and mean macroalgal index, for 27 sites (all depths combined) assessed during 1998, excluding sites with insuffieient (n<10) belt transects. (p<0.001) when all 98 sites were combined. A general inverse pattern between the piscivore density and herbivores was observed within shallow habitats. A weakly significant positive relationship existed between herbivore and piscivore density within only deep sites (p=0.034). DISCUSSION Species richness and relative family and species dominance documented during our surveys are similar to those reported in other western Atlantic reefs (Tumigan and Acosta, 1989; also see www.reef.org). However, fish densities along Andros Island are lower, particularly on fore reefs, than in some other areas of the Bahamas (Sluka et ah, 1996) and Caribbean (Schmitt, unpublished data; Lewis and Wainwright, 1985; Kramer, this volume). Given the presence and extent of well-developed reef-crest and fore-reef habitats and the relatively modest fishing pressures, a greater abundance of fishes was expected. The lack of strong spatial differences in community structure or abundance among the four geographic areas suggest that the processes governing the structure of the Andros fish populations are fairly uniform along the entire eastern reef tract. Low fish abundance in the fore reefs may be a result of limited larval supply, unsuccessful pre- and post-settlement processes, or lack of nearby juvenile habitat. No scientific surveys have been conducted on larval abundances, recruitment patterns, nursery habitat, and current patterns along Andros. The majority of larval recruitment is likely to be from local sources since the shallow, extensive carbonate platform banks surrounding the Tongue of the Ocean probably form natural barriers restricting entry of external larvae. Flow through Providence Channel is relatively low (Busby et al., 1966) and may not play a significant role in recruitment dynamics. Thus, fish populations may be influenced primarily by local recruitment processes rather than by input from external Ill larval sources (Cowen et al, 2000) and local gyres, or eddies may play a significant role in their abundance. In addition to larval supply, low fish abundances on Andros may be influenced by the availability of adjacent juvenile habitat. Nagelkerken et al. (2001) found that proximity of nearby mangroves and seagrass influenced the abundance of Ocyurus chrysurus and Scarus croicensis in Cura9ao, while the presence of available seagrass nursery habitat affected Spans oma chrysopterum, Sphyraena barracuda, and several species of Lutjanus and Haemulon. Along the eastern coast of Andros, dense seagrass beds are uncommon (total area <6 km^) (Kramer, unpublished data) and their paucity may be a limiting factor affecting fish abundance. However, the numerous creeks and bights along the mainland would appear to be ideal nursery habitats particularly as many are lined by mangroves. The relative importance of local versus external recruitment and the role of nursery habitat availability in structuring fish communities off Andros warrant further investigation. Based on our data, variance in the total density and biomass of the AGRRA fishes is mainly attributable to intrinsic habitat characteristics such as coral cover and structural complexity (Kramer et al., 1999), habitat type and depth, and sampling biases. Differences in fish assemblages between sites of similar habitat type may also result from species interaction variables (not discussed here). The significant positive relationship between chaetodontid density and live stony coral cover observed on Andros’ fore reefs may be associated with high habitat dependency or specialized microhabitat use (Robertson, 1996). The distribution of groupers has also been correlated to habitat features, particularly topographic complexity (Connell and Kingsford, 1998; Sluka et al., 1996, 2001), which may explain the relatively high abundance of M. tigris and other large-body-size groupers in the network of tunnels and overhangs between the 1 -3 m tall columns of the Montastraea annularis species complex in the fore reefs. Roberts and Ormond (1987) found the availability of shelter holes to influence the abundance of some fishes (scarids, acanthurids, labrids, and pomacentrids) while live coral cover was only important for chaetodontids. Although not examined in this study, other parameters such as adjacent habitat diversity, patch size, proximity to tidal channels, and distance from mainland may also influence fish distribution on Andros. Distinguishing which habitat factors are most important in structuring fish communities is difficult since many coexist or are additive, thus a more experimental approach isolating specific variables is needed to better understand factors governing these spatial patterns. Although acanthurids are known to prefer shallow habitats, the high abundance of scarids in reef crests off Andros is surprising since other studies have typically found their densities to be greater in deeper water (Lewis and Wainwright, 1985; also see Horn, 1989). The high abundance of acanthurids and scarids in the shallow reef crests, as well as the large size of scarids, probably result in significantly greater herbivory here than in the fore reefs. Algal communities in each habitat reflect these presumed differences in herbivory, and the strong inverse relationship between macroalgal index and herbivore biomass (Fig. 7) implies top-down control over algal assemblages. At shallow depths, the algal community is predominantly crustose corallines and turf algae, both of which are indicative of well-grazed surfaces (Steneck and Dethier, 1994). In contrast, the fore reefs 112 have few grazed surfaces and are dominated by fleshy algae (e.g., Dictyota spp., Microdictyon marinum) (Kramer et ah, this volume). Reduction of herbivory with depth has been related to decreased trophic carrying capacity (Hay and Goertemiller, 1983; Steneck, 1988). Physical factors such as sedimentation and wave energy may also contribute to algal composition in particular habitats (e.g., Fabricius and De’ath, 2001). In addition, the lack of preferred food and the increased risk of predation may also influence herbivore movements and levels of grazing. The scarcity of the important grazing sea urchin, Diadema antillarum, (Kramer et ah, this volume) is also a major factor in explaining the dominance of macroalgae in the fore reefs off Andros. Historically, Diadema was the dominant herbivore off Andros (Miner, 1933; Newell and Rigby, 1951), but populations severely declined during 1983 as in other areas of the Caribbean (e.g., Lessios et ah, 1984). Following the Diadema die- off, Morrision (1988) found that erect, resistant macroalgal species increased at intermediate depths on heavily fished Jamaican reefs. Robertson (1991) found that the abundance of two herbivorous acanthurids, Acanthurus coeruleus and A. chirurugus, increased significantly on Panama patch reefs after the die-off of Diadema. It appears that herbivorous fishes along Andros may have filled the trophic niche of Diadema as dominant grazers on high-relief reef crests but not in fore reefs. These results emphasize the importance of herbivory on Andros’ coral and algal communities. The recovery of Diadema to its former population densities may be a critical first step towards reducing macroalgae in the fore reefs. The risk of predation may influence the spatial patterns of prey fishes (e.g., Reinthal and MacIntyre, 1994), and may also partially explain the lower densities of scarids and acanthurids on Andros’ fore reefs. The high relief (2-4 m) and structurally complex arrangement of corals and pinnacles associated with intermediate depth fore-reef zones would appear to be ideal for predators because of the abundance of ambush locations. However, a weakly significant positive relationship was found between the densities of key herbivores and piscivores in the fore reefs off Andros, suggesting that other factors are also important. It is also possible that the distribution of predators determined from snapshot daytime surveys are not indicative of overall levels of predation. If fish predation on Andros’ fore reefs is unusually high, it must still be explained why densities of predators (serranids, lutjanids) are low compared to other Bahamian island groups (Sluka et al., 1996). Our survey protocol may have systematically underestimated their abundance as transect methods have been found to underreport the density of fish species with wide ranges or low abundances (Thresher and Gunn, 1986). Furthermore, the size and number of transects can also greatly influence density estimates (Sale and Sharp, 1983). Significantly fewer fish were recorded in nearly all families (including serranids) in the 1997 surveys compared to the 1998 surveys which is thought to be a direct result of the low number of transects employed at each site. However, temporal and spatial differences in the fish community for the two sample periods cannot be ruled out as additional factors. Relief and structural complexity of the habitat, by providing hiding places, may also directly influence reported fish abundances for predatory and sedentary species. It is possible that for both survey years rare and cryptic species off Andros, including most groupers, may have been systematically undersampled in fore reefs because of the 113 unusually high relief (2-4 m) of the columnar corals. A pilot comparison between 30 m fish transects swum in 7 minutes (standard) versus 1 5 minutes (extended) revealed that significantly greater numbers of serranids were seen when observers had more time to look beneath overhangs and into tunnels. Thus, observed densities of groupers may be related to the structural complexity of the substratum; the reason for the low abundance of other ecologically important (e.g., herbivores, corallivores) or commercially significant fish (e.g., snappers) in fore reefs remains unexplained. Low fish abundance, particularly of commercially significant species, on Andros fore reefs may be an indication that the entire reef tract is overfished (Roberts, 1995). Fishing pressures on Andros are thought to be light-to-moderate in comparison to many other areas of the Caribbean and are mainly targeted towards large-bodied groupers (e.g., M tigris, E. striatus and M. interstitialis) and snappers (e.g., Lutjanis analis, L. synagris and O. chiysurus). Commercial and subsistence fishers also target species such as barracuda, triggerfish, hogfish, and some grunts. Although fish such as angelfish, parrotfish, and surgeonfish are not yet targeted, there may be significant bycatch associated with trap fishing that extends to lower trophic levels (T. Turnbull, personal observation). The result of targeted fishing often leads to a decrease in the size and abundance of harvested species or, in more severe cases, to the loss of those species (e.g., Roberts, 1995; Koslow et al., 1988). Were the Andros reef tract as overfished as other reefs in the Caribbean, we might expect targeted guilds such as serranids to be dominated by species having small adult body sizes {Epinephelus fulvus, E. cruentatus and E. guttatus). In fact, large-bodied targeted species (M. tigris and E. striatus) were two of the most frequently seen serranids on Andros and nearly 30% of all groupers that were observed within transects were very large (>40 cm length). In addition, Andros is one of the few locations in the Bahamas to have at least two well-documented grouper spawning aggregations (at High Cay and Tinker Rocks, respectively). These aggregations are estimated to contain hundreds to thousands of fishes during spawning periods, although historic numbers are suspected to have been much higher (T. Turnbull, personal observations). Commercial fishing intensity has increased in the last several years, driven by higher market prices (Bahamas Reef Environmental Education Foundation and Macalister Elliot and Partners, 1998, unpublished report), particularly in southern Andros where illegal fishing by non-Bahamian residents occurs. Given the low abundance of fishes and the growing demand for targeted fish species, its fish populations are likely to be vulnerable to even modest increases in fishing intensity. Several large marine protected areas have been proposed along Andros, and at least two are in the process of being implemented (G. Larson, personal communication). In addition, in 1999 and 2000, the Bahamas Fisheries Department prohibited fishing at the High Cay aggregation site during critical spawning periods (five days around the full moon in December-February) in an effort to reduce fishing pressures. However, the aggregation was not closed in 2001 . The implementation of adaptive management strategies, such as a seasonal closure on the grouper fishery, is essential for maintaining sustainable fish populations along Andros. 114 ACKNOWLEDGMENTS This project was funded by a grant from the Committee for Research and Exploration of the National Geographic Society (Grant #6046-970). The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume as did the Hachette Filipacchi Foundation and Carolina Skiffs. Special thanks to W. Bryant, M. McGarry, and S. Cawthon for their adept piloting skills. The late Captain R. Gaensslen generously provided the use of his vessel. The Captain 's Lady, in 1997. Go Native Yacht Charters is thanked for partial charter sponsorship of their 60' catamaran in 1998. E. Fisher assisted with data entry; J. Lang and P. Richards Kramer provided many useful comments on this manuscript. We are grateful to the Bahamas Department of Fisheries for granting permission to conduct Andros fieldwork. REFERENCES Ault, J.S., S.G. Smith, J. Luo, G.A. Meesters, J.A. Bohnsack, and S.L Miller 2001 . Baseline multispecies coral reef fish stock assessments for the Dry Tortugas. Final Report, Univ. Miami, RSMAS, Miami, FL 123 pp. Bell, J.D., and R. Galzin 1 984. Influence of live coral cover on coral reef fish communities. Marine Ecology Progress Series 15:265-274. Busby, R.F., C.V. Bright, and A. Pruna 1 966. Ocean bottom reconnaissance off the east coast of Andros Island, Bahamas. Technical Report 189. US. Naval Oceanographic Office, Washington, DC. 58 pp. Connell, S.D., and M.J. Kingsford 1998. Spatial, temporal and habitat-related variation in the abundance of large predatory fish at One Tree Reef, Australia. Coral Reefs 17:49-57. Cowen, R.K., K.M.M. Lwiza, S. Sponaugle, C.B. Paris, and D.B. Olson 2000. Connectivity of marine populations: Open or closed? Science 287:857-859. Done, T.J., and R.E. Reichelt 1998 Integrated coastal zone and fisheries ecosystem management: generic goals and performance indices. Ecological Applications 8:1 10-1 18. Fabricius, K., and G. De’ath 2001 . Environmental factors associated with the spatial distribution of crustose coralline algae on the Great Barrier Reef. Coral Reefs 19: 303-309. Hay, M.E., and T. Goertemiller 1983. Between habitat differences in herbivore impact on Caribbean coral reefs. Pp. 97-102. In: M.L. Reaka (ed.). The Ecology of Deep and Shallow Coral Reefs Vol. L Symposia Series for Undersea Research, NOAA Undersea Research Program, Rockville, Maryland. 115 Horn, M.H. 1989. Biology of marine herbivorous fishes. Oceanography and Marine Biology Annual Reviews 27:167-272. Humann, P. 1 994. Reef Fish Identification. Florida Caribbean, Bahamas. 2"^* Edition. New World Publications. 396 pp. Koslow, J.A., F. Hanley, and R. Wicklund 1988. Effects of fishing on reef fish communities at Pedro Bank and Port Royal Cays, Jamaica. Marine Ecology Progress Series 43:201-212. Kramer, P.A., P.R. Kramer, and R.N. Ginsburg 1999. Spatial patterns of reef habitats and linkages to fish populations on Andros Island, Bahamas. Abstracts of 52”^ Annual Meeting of the Gulf and Caribbean Fisheries Institute, November 1-5, 1999. Key West, Florida. Lawson, G.L., D.L. Kramer, and W. Hunte 1999. Size-related habitat use and schooling behavior in two species of surgeonfish (Acanthurus bahianus and A. coeruleus) on a fringing reef in Barbados, West Indies. Environmental Biology of Fishes 54:19-33. Lessios, H.A., D.R. Robertson, and J.D. Cubit 1984. Spread of Diadema mass mortality through the Caribbean. Science 35:335-337. Lewis, S.M., and P.C. Wainwright 1985. Herbivore abundance and grazing intensity on a Caribbean coral reef. Journal of Experimental Marine Biology and Ecology 87:215-228. Lindeman K.C., G.A. Diaz, J.E. Serafy, and J.S. Ault 1998. A spatial framework for assessing cross-shelf habitat use among newly settled grunts and snappers. Gulf and Caribbean Fisheries Institute 50:385-416. McGehee, M.A. 1994. Correspondence between assemblages of coral reef fishes and gradients of water motion, depth, and substrate size off Puerto Rico. Marine Ecology Progress Series 105:243-255. Mejia, L.S., and J. Garz6n-Ferreira 2000. Reef fish community structure in four atolls of the San Andres Providencia archipelago (southwestern Caribbean). Revista de Biologia Tropical 48:883-896. Miner, R.W. 1933. Diving in coral gardens. Natural History 33:461-476. Morrison, D. 1988. Comparing fish and urchin grazing in shallow and deeper coral reef algal communities. Ecology 69:1367-1382. Nagelkerken, I., S. Kleijnen, T. Klop, R.A.C.J. van den Brand, E.C. de la Moriniere, and G. van der Velde 2001. Dependence of Caribbean reef fishes on mangroves and seagrass beds as nursery habitats: a comparison of fish faunas between bays with and without mangroves/seagrass beds. Marine Ecology Progress Series 214:225-235. Newell, N.D., and J.K. Rigby 1951. Shoal-water geology and environs, eastern Andros Island, Bahamas. Bulletin of the American Museum of Natural History 97:1-30. 116 Nunez-Lara, E., and J.E. Arias-Gonzalez 1 998. The relationship between reef fish community structure and environmental variables in the southern Mexican Caribbean. Journal of Fish Biology 53(A); 209-221. Reinthal, P.N., and I.G. MacIntyre 1994. Spatial and temporal variations in grazing pressure by herbivorous fishes: Tobacco Reef, Belize. Atoll Research Bulletin 425:1-1 1. Roberts, C.M. 1995. Effects of fishing on the ecosystem structure of coral reefs. Conservation Biology 9:988-995. Roberts, C.M., and R.F.G. Ormond 1 987. Habitat complexity and coral reef fish diversity and abundance on Red Sea fringing reefs. Marine Ecology Progress Series 41:1-8. Robertson, D.R. 1991. Increases in surgeonfish populations after mass mortality of the sea urchin Diadema antillarum in Panama indicate food limitation. Marine Biology 111: 437-444. Robertson, D.R. 1996. Interspecific competition controls abundance and habitat use of territorial Caribbean damselfishes. Ecology 77:885-899. Sale, P.F., and B.J. Sharp 1983. Correction for bias in visual transect censuses of coral reef fishes. Coral Reefs 2:37-42. Sluka, R.D., M. Chiappone, and K.M.S. Sealey 2001 . Influence of habitat on grouper abundance in the Florida Keys, USA. Journal of Fish Biology 58:682-700. Sluka, R.D., M. Chiappone, K.M. Sullivan, and R. Wright 1996. Habitat and Life in the Exuma Cays, Bahamas. Status of Groupers and Coral Reefs in the Northern Cays. Supplemental Volume: Appendices. The Nature Conservancy, Florida and Caribbean Marine Conservation Science Center. Coral Gables, FL. 216 pp. Steneck, R.S. 1 988. Herbivory on coral reefs: a synthesis. Proceedings of the Sixth International Coral Reef Symposium, Australia 1 :37-49. Steneck, R.S., and M. Dethier 1994. A functional-group approach to the structure of algal-dominated communities. Oikos 69:476-498. Thresher, R.E., and J.S. Gunn 1986. Comparative analysis of visual census techniques for highly mobile, reef- associated predators (Carangidae). Environmental Biology of Fishes 17:93-1 16. Tumigan, R.G. and A.B. Acosta 1 989. An analysis of the fish assemblages on a coral patch reef on Puerto Rico. Proceedings of the 43^^^ Gulf and Caribbean Fisheries Institute, November 1990, Miami, Florida, pp. 245-258. 117 118 119 C/D c Q < O O O O O © in 0 (X 0 cd -0 (U SP JS t, 0 C3 .s j5 Ui c3 CJ) 3 0 z. :2 u Co 00 U 00 > - m -Ss o o ^ O feb feb ^ I ® g g £ .Sf oa ■S) Z s < s s s c ^ -V O «3 g W) 60 ■£ c C J; o o ^ CJ -J <: North Grassy S26 1997 7±1 8 ±7 — 10.5 ±9.5 0 0.5 ±0.5 27 ±13 Pigeon S28 1997 5.5 ±3.5 6 ±4.5 — 0 0 0 13.5 ±7.5 South Andros* 9.9 ±5 11.5 ±6 3.3 ± 1 16.0 ±23 5.3 ± 7 0.2 ± 0 52.9 ±35 i ~ 2'' ^'iii -1--1I Mean ± standard error ; Epinephelus spp. and Myceteroperca spp. Table 2B. Density (mean ± sd) of AGRRA Fishes by site in fore reefs off Andros Island (1997 sites are italicized). Reef crest Site Year Herbivores (#/100m~) Carnivores (#/100m~) Total AGRR site name code Acanthuridae Scaridae Microspathodon Haemulidae Lutjanidae Serranidae" fishes (>5 cm) chtysurus (>5 cm) (#/100m^) 120 *0 ^ o ■H 4^ fN Os O ‘o — rsi — -H ‘O N iri ’’t -H -H r* -H ‘n ‘n •n VO 5 61 'CQ u CQ 60 c o J U S' 60 X I' V, 2 2 S o < Co 'Mean ± standard error ; ^Ep'mephelus spp. and Myceteroperca spp. 121 Table 3A. Twenty-five most frequently sighted fish species during roving diver surveys for all reef-crest sites combined off Andros Island, with density (mean ± sd) for species counted in belt transects. Scientific name Common name Roving Diver Sighting Density Frequency (SF) (1-4) SFx Density Belt transect 1998 density (#/100mh Reef-crest sites Thalassoma bifasciatum Bluehead 100 3.5 350 Acanthurus coeruleus Blue Tang 100 3.3 330 7.2 ±6.7 Abudefduf saxatilis Sergeant Major 100 3 300 — Halichoeres garnoti Yellowhead Wrasse 100 2.8 280 Acanthurus bahianus Ocean Surgeonfish 100 2.8 280 3.6 ±4.2 Lutjanus apodus Schoolmaster 100 2.7 270 3.5 ±7.4 Aulostomus maculatus Trumpetfish 96 1.9 182 Sparisoma viride Stoplight Parrotfish 96 3 287 4.7 ±5.4 Microspathodon chrysurus Yellowtail Damselfish 96 2.9 278 1.9±2.7 Scarus vetula Queen Parrotfish 96 2.8 268 2.4 ±3.2 Haemulon sciurus Bluestriped Grunt 96 2.5 240 2.4 ±4.9 Ophioblennius atlanticus Redlip Blenny 96 2.5 240 Scarus croicensis Striped Parrotfish 92 3.2 293 6.4 ±9.0 Sparisoma aurofrenatum Redband Parrotfish 92 2.9 266 1.5 ±2.1 Sparisoma rubripinne Redfm Parrotfish 92 2.8 256 1.2±2.2 Ocyurus chrysurus Yellowtail Snapper 92 2.6 238 0.9 ±3.0 Haemulon plumieri White Grunt 92 2.5 229 0.7 ± 1.6 Haemulon flavolineatum French Grunt 88 3.2 280 21.0 ±49.9 Chromis cyanea Blue Chromis 88 3.1 271 .... Stegastes partitus Biolor Damselfish 88 2.8 245 .... 122 Table 3B. Twenty-five most frequently sighted fish species during roving diver surveys for all fore-reef sites combined off Andros Island, with density (mean ± sd) for species counted in belt transects. Scientific name Common name Roving Diver Sighting Density Frequency (SF) (1-4) SFx Density Belt transect 1998 density (#/100mh Fore-reef sites Acanthurus coeruleus Blue Tang 97 3 290 3.5 ±4.0 Sparisoma viride Stoplight Parrotfish 97 2.9 280 1 .6 ± 2.6 Thalassoma bifasciatum Bluehead 94 3.4 318 .... Chromis cyanea Blue Chromis 94 3.1 290 .... Scarus croicensis Striped Parrotfish 94 3 281 5.6 ±7.3 Stegastes partitus Bicolor Damselfish 94 3 281 .... Aulostomus maculatus Trumpetfish 94 1.9 178 .... Gramma loreto Fairy Basslet 90 3 271 Sparisoma aurofrenatum Redband Parrotfish 90 2.7 244 1.0± 1.7 Ocyurus chrysurus Yellowtail Snapper 90 2.5 226 1.1 ±3.4 Canthigaster rostrata Sharpnose Puffer 90 2.4 217 .... Caranx ruber Bar Jack 90 2.3 208 0.2 ±0.9 Lutjanus apodus Schoolmaster 90 2.3 208 0.8 ±2.8 Stegastes planifrons Threespot Damselfish 90 2.2 199 .... Mycteroperca tigris Tiger Grouper 90 1.9 172 0.2 ±0.5 Clepticus parrae Creole Wrasse 87 3.7 322 .... Coryphopterus personatus Masked Goby 87 3.5 305 .... Halichoeres garnoti Yellowhead Wrasse 87 2.7 235 .... Haemulon plumieri White Grunt 87 2.1 183 0.3 ±0.7 Chaetodon capistratus Foureye Butterflyfish 87 2.1 183 0.7 ± 1.3 123 Plate 4A. The need for large-scale comparable data on coral reef condition in the Western Atlantic led to the initiation of the AGRRA Program and the development of its key reef health indicators. Given their importance in constructing the three-dimensional framework of coral reefs, the condition of scleractinian and hydrozoan corals, like this Montastraea annularis, is a primary focus of the AGRRA benthos protocol. (Photo Robert S. Steneck) Plate 4B. Distinguishing colony boundaries before estimating size or partial mortality is essential. A colony is defined on the basis of common skeletal or live tissue connections and/or by polyp size and color. Two closely adjacent colonies of Diploria labyrinthiformis are recognized by a thin lip of raised skeleton and live tissues. Common skeleton at the bases of the lobes of Montastraea annularis (Plate 3A), help in the recognition of individual colonies. (Photo Robert W. Steneck) 124 Figure 1. AGRRA survey sites at San Salvador Island, Bahamas. ASSESSMENT OF CORAL REEFS OFF SAN SALVADOR ISLAND, BAHAMAS (STONY CORALS, ALGAE AND FISH POPULATIONS) BY PAULETTE M. PECKOL,* H. ALLEN CURRAN,^ BENJAMIN J. GREENSTEIN,^ EMILY Y. FLO YD, ^ and MARTHA L. ROBBART*’^ ABSTRACT During assessments at 1 1 shallow reef sites on San Salvador Island, Bahamas in June 1998 we found low prevalence of disease, bleaching, and recent partial-colony mortality among stony corals (10 cm minimum diameter). Old partial-colony mortality was >50% in Acropora palmata\ however, recent tissue losses were low and it had recruits at several sites. Total (recent + old) partial-colony mortality of the Montastraea annularis species complex exceeded 30% on leeward patch reefs and back reefs. Groupers (serranids), snappers (lutjanids), and grunts (haemulids) were rare. Parrotfishes (scarids) were uncommon at most sites and surgeonfishes (acanthurids) were the dominant herbivores. Macroalgae, particularly browns that are seldom grazed by surgeonfishes, were the dominant algal functional group. The green macroalga Microdictyon marinum was extremely abundant and overgrowing Porites porites on leeward patch reefs. To facilitate their conservation, San Salvador Island’s reef resources should be designated as a marine reserve. INTRODUCTION The condition of coral reefs worldwide is in decline (Wilkinson, 2000) and the reefs of the greater Caribbean region have emerged as one area of particular concern (Ginsburg, 1994). Reductions in coral cover and diversity, with concomitant increases in macroalgal abundances in the Western Atlantic, have been related to the well- documented mass mortality of the herbivorous sea urchin Diadema antillarum 'Department of Biological Sciences, Smith College, Northampton, MA 01063. Email: ppeckol@email.smith.edu •y Department of Geology, Smith College, Northampton, MA 01063. Email: acurran@email.smith.edu Department of Geology, Cornell College, Mt. Vernon, lA 52314. Email : bgreenstein@comell-iowa.edu Department of Biology, San Diego State University, San Diego, CA 92182. ^ Caribbean Coral Reef Ecosystem Program, Smithsonian Institution, Washington, DC 20560. 126 (Lessios,1988; Hughes et al., 1999), to increases in coral diseases (Bruckner and Bruckner, 1997; Santavy and Peters, 1997), to overfishing (Hughes, 1994; Koslow et ah, 1 994), and to widespread elevated sea surface temperatures resulting in coral bleaching (Brown, 1997; Wilkinson et al., 1999). While the greatest degradation of Caribbean reefs is associated with large human populations (Hughes, 1994; Causey et al., 2000), fewer studies document the condition of coral reefs in areas experiencing less human disturbance as has been until quite recently the case off San Salvador, Bahamas. San Salvador Island (24®N, 74°30’W), located 600 km ESE of Miami on an isolated carbonate platform (Fig. 1), is bordered by a narrow shelf with an abrupt shelf- edge break leading to a very steep slope. Marine water quality is excellent with deep water close offshore and no immediate sources of concentrated pollutants. Its eastern and southeastern coasts typically are windward to the prevailing trade winds. A well- developed, Acropora />fl/mfl/a-dominated, bank-barrier reef lies off the northern coast and smaller bank-barrier reefs occur along the southeast and southern coasts. Hundreds of small patch reefs dot the island’s eastern shelf; larger patch reefs occur in the broad coastal embayments on the leeward western shelf. Relatively little has been published on the ecological health and short-term changes to San Salvador’s coral reefs. Curran et al. (1994) assessed stony coral cover and diversity at two leeward patch reefs between 1984 and 1992. At Telephone Pole Reef, rapidly growing colonies of Porites porites were replacing dead and broken branches of Acropora cervicornis which had been a spatial dominant until decimated by white-band disease (Aronson and Precht, 1997) during the early 1980s. During the same interval, the percent of live stony coral cover on nearby Snapshot Reef exhibited no significant change; however, there was an overall increase in the sizes of monitored coral heads and noticeably less macroalgae than seen at present. Similarly, Meyer et al. (1991) reported slight increases in the population densities of two species of the crinoid Nemaster residing in large colonies of the Montastraea annularis species complex at Snapshot Reef. More recently, stony coral cover, seaward of Snapshot Reef at the 10m CARICOMP monitoring site, experienced a slight decline and macroalgae had a twofold increase between 1994 and 1998 (Woodley et al., 1997, 2000; Gerace et al., 1998). In the Bahamas, the major commercial fishery is the spiny lobster, Panulirus argus, followed by snappers and groupers. Reports of overfishing are widespread (Woodley et al., 2000). The queen conch {Strombus gigas) provides another important, but smaller-scale (usually subsistence level), fishery although densities are declining (Stoner, 1996). Fishing regulations now prohibit the taking of nonlipped queen conch and the use of scuba gear for any kind of fishing. However, few data exist regarding the status of the marine fisheries off San Salvador Island. Until recently San Salvador has experienced little pressure from human activities. Now the island is rapidly becoming a more popular tourist destination with impetus from the opening of a large Club Med in October 1992 and recent expansion of airport facilities, including extension of the runway to permit landing of intercontinental jet aircraft. Currently, several dive boats operate primarily off the western (leeward) coast bringing up to 200 snorkel and scuba divers per day to the reefs (Kevin Collin, personal communication). With San Salvador in a state of flux with respect to human impact, the Atlantic and Gulf Rapid Reef Assessment (AGRRA) surveys reported here provide an 127 important baseline for the current condition of San Salvador’s coral reefs as well as the status of its finfish populations. METHODS Our assessment of reef-building corals, algae, and fish populations was focused on three areas of San Salvador Island (Fig. 1). Three shallow (1-4 m) fore-reef sites on the exposed northern bank-barrier at Gaulin’s Reef have low (<3 m) spur formations extending seaward from the reef crest to a carbonate pavement dominated by sea fans and Millepora spp. {Acropora palmata is the dominant coral of the reef crest.) Three sites at 2-8 m on the Gaulin’s back reef are characterized by large (1-4 m tall) colonies of the Montastraea annularis complex along with Millepora spp. and A. palmata. Three leeward patch reefs in Fernandez Bay (Snapshot, Telephone Pole, Lindsay) are each in water depths of 3-7 m. Snapshot Reef (-200 m offshore) consists of an aggregation of individual coral colonies dominated by the M. annularis complex. Telephone Pole Reef (~ 250 m offshore) is dominated by large colonies of the M. annularis complex interspersed with Porites porites growing on dead A. cervicornis. Lindsay Reef extends out from a sandy beach and experiences a relatively high sediment load. Two windward patch reefs, at depths of approximately 3-5 m and shoreward of a well-developed reef crest in French Bay, are dominated by dead A. palmata and Agaricia agaricites. These 1 1 survey sites were selected to be representative of the majority of reef types and exposure conditions occurring off San Salvador. We also considered Telephone Pole and Snapshot Reefs to be “strategic sites” because of earlier survey data (Curran et al., 1994) and their popularity as tourist dive sites. Stony coral and algal populations were assayed during June 1998 by five-six divers/survey. AGRRA Version 1 benthic protocols (see Appendix One, this volume) were used with the following modifications: stony corals >10 cm in diameter were included in the surveys; coral diameter and height were measured to the nearest cm for smaller corals (10-25 cm in diameter) and to the nearest 5 cm for larger (>25 cm) colonies. The Montastraea annularis complex was treated as a single species. Sediment deposits in the algal quadrats were removed by hand before estimating the abundance of crustose coralline algae. Diadema antillarum, being rare, was not counted. Training sessions were conducted with all divers censusing “practice” transects at the leeward patch reefs; species identifications, percent cover estimates, and coral disease and bleaching assessments were compared to ensure sampling consistency. We used Humann’s (1993) reef coral guide for most coral species identifications. A stationary visual census technique (Bohnsack and Bannerot, 1986) was employed by two divers to survey the fish populations. All sampling occurred between 10:00 a.m. and 3:00 p.m. At each sampling point, all species belonging to eight families (Acanthuridae, Chaetodontidae, Haemulidae, Labridae, Lutjanidae, Pomacentridae, Scaridae, Serranidae) observed in five minutes within a 7.5 m radius cylinder were recorded. Each census was begun three minutes after laying a measuring tape on the substratum by counting all individuals of all species observed in the pre-set radius within the initial field of view. New sectors of fields of view were then scanned by rotating in one direction. Abundances of species moving in schools were taken when first observed 128 in the sampling cylinder (it was important to count fishes moving in schools immediately because they were unlikely to remain in the sampling area). When very large schools were present it was sometimes necessary to estimate numbers in 10s or 50s. Fish lengths were estimated in cm using a T-shaped tool marked every 5 cm to help avoid underwater magnification problems (Bohnsack and Bannerot, 1986). We recorded the number of individuals, plus the minimum, maximum, and mean estimated lengths for each of the eight fish families. We used Humann’s (1994) reef fish guide for species identifications. Littler and Littler (2000) was later consulted for macroalgae. As time permitted on the leeward patch reefs, two divers measured herbivorous fish grazing rates following the AGRRA methodology given in Appendix One. All grazing rate measurements were made between 10:00 a.m. and 2:00 p.m. during the peak time for grazing activity (Lewis, 1986). RESULTS Stony Corals Species composition. At each site, we censused at least 1 0 transects with each transect sampling ~9-12 stony corals that were >10 cm in diameter (Table 1). In terms of numerical abundance, Acropora palmata was fairly important (means of 13-19%) along the Gaulin’s bank-barrier reef and on the windward patch reefs at French Bay (Fig. 2A, B,D). While >50% of the upper surfaces of the A. palmata colonies were long dead (means for the six Gaulin’s sites ranged from 42-78%; mean of 65% at French Bay), very little (<2%) recent partial mortality of colony surfaces (hereafter recent mortality) was evident for this species. The Montastraea annularis complex dominated (mean 40% of all colonies) the leeward patch reefs in Fernandez Bay (Fig. 2C). Recent mortality here was fairly low (<4% of their upper surfaces), while nearly 40% (range 30-48% for three leeward patch reefs) of the corresponding surfaces were long dead. The Montastraea annularis complex and Diploria clivosa together contributed about 25% of the colonies in Gaulin’s back reef (Fig. 2B). Approximately 35% of the upper surfaces of colonies of M annularis complex on the back reef were long dead. Millepora complanata, which forms extensive thickets on the spurs, contributed nearly a quarter (24%) to the total abundance of stony corals at Gaulin’s fore reef (Fig. 2A). Colonies of Agaricia spp. were important on the windward patch reefs and Porites spp. were numerically abundant at most sites (Fig. 2). Recruits. The composition of recruits (Fig. 3) was largely dominated by Porites astreoides (especially on the bank-barrier reef), P. porites (all patch reefs, Gaulin’s back reef) and Agaricia agaricites (especially on windward patch reefs). Collectively, poritids and agariciids contributed 45-70% of the recruit densities. Although 15% of the recruits were of the Montastraea annularis complex on the leeward patch reefs and accounted for <10% of the recruits on the bank-barrier reef, none were found on the windward patch reefs. At all four habitat types, Acropora palmata represented <5% of the recruits. 129 A Bank-barrier Fore B Bank-barrier Back C Leeward Patch Reefs D Windward Patch Other 1 0% Other 27% PP 14% AG 7% PP 4% IVlA40% PA 11% AP 19% ’//////////? V////////////\ ■//////////////>\ ///////////////X y/////////////. <• ‘/////////.'yr. ///// MA 7% ,?lk ' AGT 8% MILC 6% PA 12% DC 5% ^ AP=Acropora palmata MA=Mo/ifasfraea annu/aris complex AG=Agaricia agaricites ^ AGT= Agaricia tenuifolia ^ DC=Diploria clivosa ^ MILC=/M;7/epora complanata PA=Porites astreoides PP=Porites porites Other Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>10 cm diameter) at (A) bank-barrier fore reef (n=319), (B) bank-barrier back reef (n=232), (C) leeward patch reefs (n=344), (D) windward patch reefs (n=249), off San Salvador Island, Bahamas. Other category = combined coral species, each with <5% abundance of occurrence. 130 A Bank-barrier Fore Reef B Bank-barrier Back Reef AC 6% PA 33% C Leeward Patch Reefs D Windward Patch Reefs AC=Acropora cervicornis AG=Agaricia agaricites AGJ=Agaricia tenuifolia □ DC=Diploria clivosa EF=Eusmilia fastigiata MILC=/W///eDora complanata MA=/Wonfasfraea annularis complex PA=Porites astreoides PP=Porites porites SS=Siderastrea siderea Other Figure 3. Species composition and mean relative abundance of all stony coral recruits (<2 cm diameter) at (A) bank-barrier fore reef (n=64), (B) bank-barrier back reef (n=52), (C) leeward patch reefs (n=96), (D) windward patch reefs (n=65), off San Salvador Island, Bahamas. 131 Figure 4. Size-frequency distribution as % of dominant stony corals (>10 cm diameter) at (A) bank- barrier fore reef, (B) bank-barrier back reef, (C) leeward patch reefs, (D) windward patch reefs, off San Salvador Island, Bahamas. 132 Percentage of Coral Colony Dead Figure 5. Frequency distribution of total (recent + old) partial colony mortality of all stony corals (>10 cm diameter) at (A) bank-barrier fore reef, (B) bank-barrier back reef, (C) leeward patch reefs, (D) windward patch reefs, off San Salvador Island, Bahamas. 133 A Bank-barrier Fore Reef B Bank-barrier Back Reef (D D 0) (D C3 C3 C3 a 03 03 03 “O 72 72 '!2 ;o c 'u 'u *c: C D 03 o ■o E 03 D C O o O CA) u u o X 03 on "5 C3 -r: U X s o eu Fish Families C Leeward Patch Reef D Windward Patch Reef Fish Families Fish Families Figure 6. Mean fish abundance (no. individuals/200 m^) at (A) bank-barrier fore reef, (B) bank-barrier back reef, (C) leeward patch reefs, (D) windward patch reefs, off San Salvador Island, Bahamas. 134 Coral size. The mean size of surveyed corals (>10 cm in diameter) ranged from 21 cm (at Lindsay Reef) to 85 cm (at Gaulin’s A). However, mean colony sizes were remarkably similar among habitat types ranging from ~40 cm on the leeward patch reefs to ~53 cm in the bank-barrier back reef (Table 2). Size-frequency distributions of the major reef-building stony corals at each habitat type are shown in Figure 4. Wherever Acorpora palmata occurred a high proportion of colonies were very large (>200 cm), emphasizing its importance as a major frame-builder. Millepora complanata also showed a relatively broad size range with colonies of <20 cm to >200 cm in diameter at the bank- barrier fore reef While the Montastraea annularis complex only exceeded 1 00 cm in diameter on the leeward patch reefs, its massive growth forms provided substantial relief at most sites. Coral condition. We observed evidence of coral disease in 2% or less of the censused colonies at most (8/1 1) sites (Table 2). Signs of disease were higher at one windward (3%) and two of the leeward patch reefs (4% at Telephone Pole; 8% at Snapshot). The coral diseases observed were black-band disease and yellow-blotch disease (both mainly on the M. annularis complex) and white-band disease (on^. palmata). The percentage of colonies that were bleached was low in all habitat types (from 1.3% in Fernandez Bay to 2.6% in Gaulin’s fore and back reefs). Recent mortality was also very low at most sites (Table 2). The higher percentages of recent mortality at Snapshot (4%) and Telephone Pole (4.7%) were largely associated with the Montastraea annularis complex and Porites porites, respectively. The decline of P. porites was due in large part to overgrowth by the macroalga Microdictyon marinum and crustose coralline algae. Although old partial-colony mortality (hereafter old mortality) exceeded 30% at three sites (Gaulin’s back reef A, Snapshot, Telephone Pole), it largely was confined to the locally dominant coral species (Table 2). Total (recent + old) partial-colony mortality (hereafter total mortality) was lowest overall in the bank-barrier fore reef (-20%) and highest at Snapshot (-35%, where total mortality of the M. annularis complex was nearly 55%), Gaulin’s A (-38%, where A palmata and the M annularis complex together contributed 65% of the colonies measured) and Telephone Pole (-42%, with P. porites exhibiting about 44% total mortality). Because recent mortality was never greater than 5% of the upper surface of coral colonies, we plotted frequency distributions of total mortality for the four reef habitat types (Fig. 5A-D). In all areas the vast majority of corals had <10% total mortality. However, there was a clear signature (-5-10% of stony corals measured) of colonies showing >90% total mortality at all sites except the windward patch reefs. Algae Macroalgae dominated the benthic algae in three habitats (Gaulin’s fore and back reefs, leeward patch reefs) and was the predominant component of the algal assemblages at all but three sites (Table 3). On leeward patch reefs, where their abundance was highest (about 50 to 70%), the fleshy green Microdictyon marinum was the dominant species. For example, at Telephone Pole Reef it commonly was completely overgrowing Porites porites resulting in total colony mortality. Elsewhere, brown algae, including Dictyota 135 divahcata, D. bartayresii, Lobophora variegata, Padina sanctae-crucis, Turbinaria turbinata and Stypopodium zonale, predominated. Green calcareous Halimeda spp. were extremely rare at all reef sites. Mean macroalgal height was generally lowest (~2 cm) at the windward sites (fore reef and patch reefs) but reached nearly 3 cm at Telephone Pole Reef. Mean macroalgal indices (absolute abundance of macroalgae x macroalgal height) varied from about 43 on the windward patch reefs to 150 on the leeward patches. Turf algae were more abundant at two sites (Gaulin’s fore reef 1, French Bay 2) while crustose coralline algae were predominant at only a single site (French Bay 1). Fishes Mean fish densities (Table 1) were highest at the leeward patch reefs, ranging from ~50 (Lindsay Reef) to 177 individuals/ 200 m^ (Snapshot Reef). Fish abundances were relatively similar at the other sites varying between ~43 (windward patch reefs and bank-barrier back reef) and 53 individuals /200 m^ (bank-barrier fore reef). All reef habitats showed a dominance of wrasses (Labridae) and surgeonfishes (Acanthuridae) which collectively represented 76% of the total fish abundance on the leeward and windward patch reefs and over 90% on the bank-barrier reefs (Fig. 6A-D). Parrotfish (Scaridae) densities were low at all sites except Lindsay Reef where acanthurids were scarce (Table 4), yet the highest abundances of surgeonfishes occurred at the two other leeward patch reefs (Snapshot and Telephone Pole). Groupers and other seabasses (Serranidae), snappers (Lutjanidae) and grunts (Haemulidae) were absent altogether on the Gaulin’s back reef, present in extremely low densities on the Gaulin’s fore reef and the windward patch reefs, and slightly more abundant on the leeward patch reefs (Fig. 6; Table 4). Considered together, these three families represented ~12% of total abundance in the eight censused fish families at Snapshot and Telephone Pole Reefs. The size-frequency distributions for two major guilds (herbivores-parrrotfishes, surgeonfishes, the yellowtail damselfish Microspathodon chrysurus; camivores-grouper, snapper) at each of the four reef types are shown in Figure 7. Notably, no groupers or snappers were censused on the bank-barrier back-reef Relatively low abundances of carnivores elsewhere make length comparisons among sites difficult. For all reef types, 40-50% of the herbivores fell within the 11-20 cm length category; generally, <10% were larger than 20 cm. We found no relationship between herbivorous fish density and macroalgal index (Fig. 8). Grazing rates were nearly identical at Telephone Pole (mean=3.1 bites/minute, se =0.9, n=8) and Snapshot (3.5 bites/minute, se=l.l, n=12) Reefs, respectively. The grazing rate at Lindsay Reef was nearly three times those values (mean=l 1.1 bites/minute, se=0.4, n=5). Ninety percent of the grazing at Lindsay Reef was due to parrotfishes, whereas surgeonfishes were the dominant herbivores at the other two patch reefs. DISCUSSION The stony corals at 1 1 evaluated sites of the reef system off San Salvador Island, Bahamas, are in reasonably good condition. Total (recent + old) partial-colony mortality 136 of upper surfaces (colonies >10 cm diameter) ranged from 15-42% and was particularly low (-20%) on the bank-barrier fore reef. Interestingly, these estimates of old (and total) mortality were lowest (-15-16.5 %) at the sites on the fore reef and back reef that are nearest to land. These two sites experience higher wave surge and mixing than elsewhere on Gaulin’s bank-barrier reef which perhaps provides a more favorable environment for coral survival. Similar to other Caribbean reefs, Acropora palmata, formerly a dominant in shallow windward zones, showed high (>50%) total colony mortality; nevertheless, we found evidence of its recruitment at several sites. As found in other studies (e.g., Rogers et al., 1986), the species composition of the stony coral recruits did not reflect the major coral-reef builders at any site. Although 15% of the recruits at the leeward patch reefs were of the Montastraea annularis complex, which here represented 40% of the surveyed (>10 cm) corals, no recruits (and B Bank-barrier Back Reef 1.00 0.80 0.60 0.40 0.20 0.00 ■ Herbivores n=49 ^Carnivores n=0 ll.- 0-5 6-10 11-20 21-30 31-40 >40 Length (cm) categories D Windward Patch Reef Length (cm) categories A Bank-barrier Fore Reef 1.00 Cl 0.80 © © 0.60 c ^© 0.40 © a. © 0.20 u a. 0.00 ■Herbivores n=42 '^Carnivores n=6 111 0-5 6-10 11-20 21-30 31-40 >40 Length (cm) categories C Leeward Patch Reef 1.00 0.80 0.60 O W o c t 0.40 o a. o 0.20 a. ■Herbivores n=44 '^Carnivores n=19 0.60- 0.40- I r 0.20' ■_ m o.ool ■ , i~i , ri J I 0-5 6-10 11-20 21-30 31-40 >40 Length (cm) categories Figure 7. Size-frequency distribution of herbivores (all acanthurids and scarids, Microspathodon chrysurus) and carnivores (all lutjanids and serranids) at (A) bank-barrier fore reef, (B) bank-barrier back reef, (C) leeward patch reefs, (D) windward patch reefs, off San Salvador Island, Bahamas. 137 few >10 cm colonies) were found on the windward patch reefs. This species complex was a fairly important component (16% of >10 cm corals) in the bank-barrier back-reef region but it was represented by only 7% of the recruits. Although receiving higher sediment loading from its proximity to the beach, recent and old mortality values were significantly lower at Lindsay Reef than on the other two leeward patch reefs. This apparent paradox may be an artifact of our sampling. Lindsay is a fairly “dead reef’ overall; however, the live corals that remain are in good condition. The high percentages of partial mortality of the Montastraea annularis complex at Snapshot Reef, possibly from disease, and of Porites porites at Telephone Pole Reef from algal overgrowth, both occurring since Curran et al.’s (1994) assessment of these reefs, are causes for concern. The M annularis complex has also suffered high rates of partial-colony mortality at other Caribbean sites. For example, two major warming events (1995, 1998) in Belize have resulted in massive coral bleaching with subsequent increased evidence of coral disease and mortality (McField, 1999; Peckol et al., this volume). Although the high incidence of diseased corals at Snapshot Reef (8% of censused colonies, 2.5% belonging to the M. annularis complex) in June 1998 predated the 1998 warming event that resulted in major coral bleaching worldwide (Wilkinson, 2000), San Salvador’s leeward patch reefs had previously bleached in 1995 (McGrath and Smith, 1999). However, these researchers noted that the major effect of the warming event was experienced by Agaricia spp. not Montastraea. By June 1998, colonies of Porites porites were no longer expanding over the skeletons of Acropora cervicornis at Telephone Pole Reef. A more recent AGRRA survey during June 2000 showed that their condition had declined even further; partial- 2001 180- 160- ■o 140- 1d 120- 2 100- 1 80- S 60- 40- ‘ • 20- Q j j , , , , 0 20 40 60 80 100 120 Mean Herbivore Density (individuals/200ni^3) Figure 8. Relationship between mean herbivore abundance (no. individuals/200 m3) and mean macroalgal index, by site in Sal Salvador. 138 colony mortality had risen from 44 % to >50% of the upper surfaces concomitant with an increase in macroalgal abundance from 57% to 88% (Peckol et ah, unpublished). In Belize, Lewis (1986) demonstrated that macroalgae can directly overgrow and kill portions of Pontes astreoides in herbivore exclusion treatments. Increases in macroalgae associated with declining coral cover have also been documented on other Caribbean reefs, including the San Bias Islands, Panama (Ogden and Ogden 1994), Jamaica (Hughes 1994), and even areas remote from human impact (McClanahan et al, 1999). High macroalgal abundances on San Salvador reefs may be related to the composition of the herbivorous fish guild where, in 1998, acanthurids (surgeonfishes) predominated at most sites. Lewis and Wainwright (1985) reported highest grazing rates for Belize in areas supporting higher parrotfish densities; similarly our highest grazing rates in Fernandez Bay were at Lindsay Reef where scarids were more common. Lewis (1985) also noted that parrotfish actively graze several genera of brown algae, including Sargassum, Turbinaha, and Padina, that are common at many of the San Salvador sites. Brown algae were not grazed at all by two surgeonfishes in Belize (Lewis 1985) and showed significant increases in percent cover and height (i.e., height is not a measure of abundance but rather of size/biomass) in response to experimental reduction in herbivory (Lewis, 1986). These findings may help to explain the high relative abundances of macroalgae off San Salvador reefs which are dominated by brown seaweeds at all but the leeward patch reefs. Lewis’ (1985, 1986) results also may explain why there was no relationship between herbivore density (primarily acanthurids) and macroalgal index in the present study. Absent or rare in most reef habitats, snappers, groupers and grunts were found in relatively high numbers only at Snapshot Reef in Fernandez Bay; this area may have a somewhat lower level of fishing activity than other sites. Hence, the San Salvador reefs are probably experiencing pressures from overfishing, but this conclusion cannot be stated with certainty because currently there are no comparable areas off San Salvador designated as “no-take” zones. In similar patch reefs off Belize, snapper and grouper densities and lengths are significantly greater within marine reserves and areas nearby (spillover effect) compared with sites not protected from fishing (Sedberry et al., 1992; Peckol et al., this volume). Although we have documented partial mortalities of the major reef-building corals, Acropora palmata and the Montastraea annularis complex, that are relatively high in some habitats, the San Salvador reef system was fairly resistant to a large-scale disturbance from the passage of Hurricane Floyd. On September 13-14, 1999, this Category 4 storm passed within 20 to 30 nautical miles NE and N of the island with winds reaching 135 knots. Its greatest impact was felt on the leeward side of the island which experienced substantial coastal erosion and damage to buildings and infrastructure. However, the leeward patch reefs in Fernandez Bay showed little damage. We resurveyed the three leeward patch reefs in January 2000 and found insignificant change from June 1 998 in the percent of recent or old partial mortality for the corals at these sites (compare Tables 2 and 5). As Smith and Buddemeier (1992) suggested for other reef systems, the San Salvador reefs displayed resilience in response to this large-scale natural disturbance. Notwithstanding San Salvador’s remote location, excellent marine water quality, and low human population density, the coral reefs surrounding the island are experiencing increased pressures from the combined effects of tourism and possibly 139 overfishing. Maintaining the integrity of San Salvador’s coral reefs and adjacent marine ecosystems, and increasing the populations of fishes and invertebrates, will be important to the island’s economic future. We strongly recommend the establishment of a marine reserve with an active management and regulations enforcement plan for all or a portion of San Salvador’s reef system. Such designation should facilitate conservation of its critical fish and coral resources (Roberts, 1995; Nowlis and Roberts, 1997) and might contribute larval and adult fish to adjacent Bahamian insular and bank areas (Russ and Alcala, 1996; Stoner, 1996). ACKNOWLEDGMENTS This research was funded by a Culpeper Foundation grant to Smith College (Curran and Peckol, coprincipal investigators), by the Smith College Summer Science Program, and by the B. Elizabeth Homer Fund. We greatly appreciate the field- and data- processing assistance provided by Smith College AGRRA team members: Lora Harris, Hali Kilboume, Michelle Kolpin, and Shannon Ristau. We thank the Bahamian Field Station for facilities and staff support. REFERENCES Aronson, R.B., and W.F. Precht 1997. Stasis, biological disturbance and community structure of a Holocene coral reef Paleobiology 23:326-346. Bohnsack, J.A., and S.P. Bannerol 1986. A stationary visual census technique for quantitatively assessing community structure of coral reef fishes. NOAA Technical Report National Fish and Wildlife Service 41:1-15. Brown, B. 1997. Coral bleaching: causes and consequences. Proceedings of the Eighth International Coral Reef Symposium, Panama City, 1:65-74. Bruckner, A., and R. Bruckner 1997. Spread of a black-band disease epizootic through the coral reef system in St. Ann’s Bay, Jamaica. Bulletin of Marine Science 61:919-928. Buddemeier, R.W. 1992. Corals, climate and conservation. Proceedings of the Seventh International Coral Reef Symposium, Guam 1:3-10. Causey, B., J. Delaney, E. Diaz, D. Dodge, J.R. Garcia, J. Higgins, W. Jaap, C.A. Matos, G.P. Schmahl, C. Rogers, M.W. Miller, and D.D. Turgeon 2000. Status of coral reefs in the U.S. Caribbean and Gulf of Mexico: Florida, Texas, Puerto Rico, U.S. Virgin Islands and Navassa. Pp. 261-285. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science. Cape Ferguson, Queensland and Dampier, Western Australia. 140 Curran, H.A., D.P. Smith, L.C. Meigs, A.E. Pufall, and M.L. Greer 1994. The health and short-term change of two coral patch reefs, Fernandez Bay, San Salvador Island, Bahamas. Pp. 147-153. In: R.N. Ginsburg (compiler) Global Aspects of Coral Reefs - Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Gerace, D.T., G.K. Ostrander, and G.W. Smith 1998. San Salvador, Bahamas. Pp. 229-245. In: B. Kjerfve (ed.), CARICOMP- Caribbean Coral Reef, Seagrass and Mangrove Sites. Coastal region and small island papers 3, UNESCO Paris. Ginsburg, R.N. (compiler) 1994. Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards, and History, Rosenthiel School of Marine and Atmospheric Science, University of Miami, Miami. 420 pp. Hughes, T.P. 1994. Catastrophes, phase shifts and large-scale degradation of a Caribbean coral reef. Science 265:1547-1550. Hughes, T.P., A.M. Szmant, R. Steneck, R. Carpenter, and S. Miller 1999. Algal blooms on coral reefs: what are the causes? Limnology and Oceanography 44:1583-1586. Humann, P. 1993. Reef Coral Identification. New World Publications, Inc., Jacksonville, FL, 252 pp. Humann, P. 1994. Reef Fish Identification, 2"*^ ed. New World Publications, Inc., Jacksonville, FL, 24 pp. Koslow, J.A.K. Aiken, S. Avil, and M.A. Clements 1994. Catch and effort analysis of the reef fisheries of Jamaica and Belize. Fisheries Bulletin. 92:737-747. Lessios, H.A. 1988. Mass mortality of Diadema antillarum in the Caribbean: what have we learned? Annual Review of Ecology and Systematics 19:371-393. Lewis, S.M. 1985. Herbivory on coral reefs: algal susceptibility to herbivorous fishes. Oecologia 65:70-375. Lewis, S.M. 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Lewis, S.M., and P.C. Wainwright 1985. Herbivore abundance and grazing on a Caribbean coral reef Journal of Experimental Marine Biology and Ecology 87:215-228. Littler, D.S., and M.M. Littler 2000. Caribbean Reef Plants. An Identification Guide to the Reef Plants of the Caribbean, Bahamas, Florida and Gulf of Mexico. Offshore Graphics, Inc., Washington, D.C. 542pp. McClanahan, T.R., R.B. Aronson, W.F. Precht, andN.A. Muthiga 1999. Fleshy algae dominate remote coral reefs of Belize. Coral Reefs 18:61-62. 141 McField, M.D. 1999. Coral response during and after mass bleaching in Belize. Bulletin of Marine Science 64:155-172. McGrath, T.A., and G.W. Smith 1999. Monitoring the 1995/1996 and 1998/1999 bleaching events on patch reefs around San Salvador Island, Bahamas. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration, Fort Lauderdale, pp 135-136. Meyer, D.L., B.J. Greenstein, and G. Llywellyn 1991 . Population stability of crinoids at Snapshot Reef, San Salvador, Bahamas. Pp. 181-184. In: R.J. Bain (ed.). Proceedings of the Fifth Symposium on the Geology of the Bahamas, Bahamian Field Station, San Salvador. Nowlis, J.S., and C.M. Roberts 1997. You can have your fish and eat it too: theoretical approaches to marine reserve design. Proceedings of the Eighth International Coral Reef Symposium, Panama City, 2:1907-1910. Ogden, J.C., and N.B. Ogden 1993. The coral reefs of the San Bias Islands: revisited after 20 years. Pp. 267-271. In: Ginsburg R.N. (compiler), Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards, and History. Rosenthiel School of Marine and Atmospheric Science, University of Miami. Roberts, C.M. 1995. Rapid build-up of fish biomass in a Caribbean marine reserve. Conservation Biology 9:815-826. Rogers, C.S., H.C. Fritz, III, M. Gilnack, J. Beets, and J. Hardin 1986. Scleractinian coral recruitment patterns at Salt River Submarine Canyon, St. Croix, U.S. Virgin Islands. Coral Reefs 3:69-76. Russ, G R., and A.C. Alcala 1996. Do marine reserves export fish biomass: evidence from Apo Island, central Philippines. Marine Ecology Progress Series 132:1-9. Santavy. D.L., and E.C. Peters 1997. Microbial pests: coral disease in the Western Atlantic. Proceedings of the Eighth International Coral Reef Symposium, Panama City 1:607-612. Sedberry, G., J. Carter, and P. Barrick 1 992. The effects of fishing and protective management on coral reefs of Belize. Proceedings of the Gulf and Caribbean Fisheries Institute 1 1 : 1-25. Smith, S.V., and R.W. Buddemeir 1 992. Global change and coral reef ecosystems. Annual Review of Ecology and Systematics 23:89-1 18. Stoner, A.W. 1996. Queen conch, Strombus gigas, in fished and unfished locations of the Bahamas: effects of a marine fishery reserve on adults, juveniles, and larval production. Fishery Bulletin 94:551-565. 142 Wilkinson, C., O. Linden, H. Cesar, G. Hodgson, J. Rubens, and A.E. Strong 1999. Eeological and socioeconomic impacts of 1998 coral mortality in the Indian Ocean: an ENSO impact and a warning of future change? Ambio 28:188-196. Wilkinson, C. 2000. Executive summary. Pp. 7-17. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science, Cape Ferguson, Queensland and Dampier, Western Australia. Woodley, J.D., P. Alcolado, T. Austin, J. Barnes, R. Claro-Madruga, G. Ebanks-Petrie, R. Estrada, F. Geraldes, A. Glasspool, F. Homer, B. Luckhurst, E. Phillips, D. Shim, R. Smith, K. Sullivan-Sealy, M. Vega, J. Ward, and J. Wiener 2000. Status of coral reefs in the northern Caribbean and western Atlantic. Pp: 261-285. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science, Cape Ferguson, Queensland and Dampier, Western Australia. Woodley, J.D., K. De Meyer, P. Bush, G. Ebanks-Petrie, J. Garz6n-Ferreira, E. Klein, L.P.J..J. Pors, and C.M. Wilson 1997. Status of coral reefs in the south central Caribbean. Proceedings of the Eighth International Coral Reef Symposium, Panama City 1:357-362. Table 1. Site information for AGRRA stony coral, algae and fish surveys off San Salvador Island, Bahamas, June 1998. 143 C/D g 03 rG cd PQ uT O 03 "Id 00 c o3 00 0-1 C3 Cl, ■T3 30% cover at six sites. Parrotfish densities exceeded surgeonfishes at most sites (1 1/13). Consistent patterns of lower partial-colony mortality of stony corals and greater fish densities and sizes near and within the Hoi Chan Marine Reserve highlight the ecological benefits of protected areas for the maintenance of reef corals and attendant fish populations. ^Department of Biological Sciences, Smith College, Northampton, MA 01063. Email: ppeckol@email.smith.edu ^Department of Geology, Smith College, Northampton, MA 01063. Email: acurran@email.smith.edu ^Department of Biology, San Diego State University, San Diego, CA, 92182. "^Caribbean Coral Reef Ecosystem Program, Smithsonian Institution, Washington, DC 20560. ^Department of Geology, Cornell College, Mt. Vernon, lA 52314. Email : bgreenstein@comell-iowa.edu 148 INTRODUCTION Over the past two decades, there has been widespread deterioration of coral reefs worldwide (Wilkinson, 2000). Following the mass mortality in 1983 of a “keystone” herbivore, the sea urchin Diadema antillarum, Caribbean reefs emerged as an area of particular concern (see Ginsburg, 1994) when macroalgal abundances increased at many reefs (Lessios, 1988; Hughes et al., 1999; but see Lapointe, 1997, 1999). The Belize barrier reef system largely was spared from the negative effects of the disappearance of D. antillarum because, prior to the mass mortality event, its densities were extremely low compared with those on reefs experiencing concurrent pressures of overfishing (Hay, 1984). Until 1995, the Belize reefs had also escaped the mass coral bleaching events reported in many other areas of the Caribbean (Macintyre and Aronson, 1997; McField, 1999). In addition, Belize has a very low human population density (about 230,000 or about nine people/km^). The central and southern regions of the Belize barrier reef are distant from the effects of terrestrial development and, at least until recently, fishing has been primarily at the subsistence level. The Belize barrier reef ecosystem thus must have served as a large and important regional source of larvae and juvenile corals, other invertebrates, and fishes (Cortes, 1997). The coupling of low incidence of large-scale natural disturbances and minimal anthropogenic effects conserved this reef system in a nearly pristine state with a unique array of reef types and luxuriant coral communities (Perkins and Carr, 1985). However, the hurricanes in the ‘60s and 70’s that Stoddart (e.g., 1974) made famous and the effects of white-band disease, all occurring before the 1995 bleaching, demonstrate that natural perturbations pre-1995 are not without their effects in Belize. The Belize barrier reef complex is the largest continuous reef system in the western North Atlantic-Caribbean Province extending a distance of 250 km south from the northern end of Ambergris Caye. It is one of the best-studied coral-reef areas in the western hemisphere (Cortes, 1997), particularly since the establishment of the Smithsonian Institution’s field station on Carrie Bow Caye in 1972 (Rutzler and Macintyre, 1982; Macintyre and Aronson, 1997). A major NNE-trending fault system is clearly reflected in the alignment of the coastline, barrier reef, and three oceanic atolls. The reef is fringing at its northernmost end, but as it extends southward the system becomes a nearly continuous barrier with well-developed fore reefs, reef crests, and extensive back-reef (lagoonal) areas. In the shallower portions of the fore-reef zone, interdigitating coral buttresses and sandy troughs form a sometimes massive spur-and- groove system. The reef crest is a high-energy zone consisting of a shallow rampart built of coral rubble and dominated by the coral Acropora palmata (James and Ginsburg, 1979). The back-reef area has a large number of cayes (cays, or small islands), many with associated patch reefs. There is a complex network of steep-sided, rhomboidal-shaped shoals and reefs (herein referred to as coral reef ridges) in the south-central lagoon, which originated in fault patterns enhanced by subsequent Pleistocene karst solution and differential carbonate deposition (Macintyre and Aronson, 1997; Macintyre et al., 2000). Coral reef ridges previously supported spectacular coral development dominated by the staghorn Acropora cervicornis at depths of 3-8 m. Following mass mortality in the 1980s of A. cervicornis from white-band disease, the ridges were rapidly colonized by Agaricia spp., primarily /I. tenuifolia (Aronson and Precht, 1997). 149 The biologically rich Belize barrier reef ecosystem contributes to a multimillion- dollar fishing industry and serves as a prime attraction for sport fishermen and scuba/snorkel divers. Tourism in Belize (see http://www.belizetourism.org/arrival.html) has been expanding rapidly from less than 100,000 visitors in 1985 to over 326,000 in 2000 and growing. Tourism generates over 26% of the Belizean gross national product (Higinio and Munt, 1993). Ambergris Caye has borne the brunt of the effects of coastal tourism including degradation of the reef, dwindling fish stocks, and surface runoff In 1987, the Government of Belize established the Hoi Chan Marine Reserve off Ambergris Caye in an effort to conserve a small but complete portion of this reef ecosystem including the coral reef, lagoon, and mangrove habitats (Carter et ah, 1994). Modeled in large part after the Australian Great Barrier Reef management plan, this reserve is zoned with a multi-use management scheme that has achieved some success. Although the Hoi Chan Marine Reserve does not encompass an expansive reef tract, it was designed to serve as an impetus for the creation of additional reserves in the future (Carter et ah, 1994). Within the past decade the Belize barrier reef has undergone a noticeable decline (Kramer et ah, 2000), particularly as fishing pressures have reduced stocks (Carter et ah 1994). Several major disturbances, including warming events during 1995/96 (Burke et ah, 1996; McField, 1999) and 1998 (Mumby, 1999; Aronson et ah, 2000), and a near direct hit by Hurricane Mitch in autumn 1998, have accelerated its degradation. The 1998 bleaching event began in September following a month of calm weather and increasing water temperatures. The eye of Hurricane Mitch, a category 5 storm, passed -200 km SE of Glovers Reef in late October 1998 (Mumby, 1999). The subsequent demise of Agaricia tenuifolia on the coral reef ridges in the south-central region, a phenomenon unprecedented in recent geologic history, has been particularly well described by Aronson et ah (2000). Given the large size and the diversity of reef types present on the Belize barrier, there is a great need for larger-scale monitoring to document the short- and long-term effects of these widespread disturbances on this very important reef system. In addition, understanding the effects of marine reserves on the health and conservation of coral and fish populations is critical for long-term management decisions. Hence, we investigated the condition of stony coral, algae and fish populations at 13 sites along the northern and south-central regions of the Belize barrier reef, including fore reefs, patch reefs and coral reef ridges. The timing (May 1999) of our survey enabled us to document the effects of the major 1998 warming event and Hurricane Mitch. The Hoi Chan Marine Reserve was included in our surveys to provide information useful for management decisions regarding this and other reserves planned for the Belize barrier reef METHODS The 13 sites that we surveyed during May 1999 were selected as being representative of the reef types and conditions in the northern and south-central regions of the Belize Barrier Reef (Fig. 1). The four northern sites near Ambergris Caye were also strategically chosen to compare areas within and outside the Hoi Chan Marine Reserve. 150 Five fore-reefs were censused. San Pedro Canyon, located south of Mexico Rocks off Ambergris Caye at 12-15 m depth, is characterized by a low-relief (~2 m) spur-and- groove topography with small colonies of the Montastraea annularis species complex and numerous gorgonian corals. Eagle Ray Canyon (depth 12-18 m), which lies north of the Hoi Chan Channel but is still within the reserve, has a well-developed spur-and- groove system. It supports a high diversity of corals in good condition dominated by large colonies of the M. annularis complex and Agaricia tenuifolia. Tobacco Reef, in south-central Belize, has a low-relief spur-and-groove formation at depths of about 12-13 m with large, healthy colonies of the M annularis complex. A. tenuifolia was overgrowing dead skeletons of Millepora complanata on the fore-reef spurs. (In May 1999, the large coral-rubble ridge that runs a considerable distance from north of Tobacco Caye southward to near South Water Caye was covered with large fragments of what appeared to be recently dead coral colonies that were reported by locals to have been deposited by Hurricane Mitch.) The fore reef off South Water Caye (up to 14 m depth) has a high-relief (3-5 m) spur-and-groove development with much Siderastrea siderea and some large colonies of the M annularis complex, Millepora complanata and Agaricia spp. were abundant on the spurs. Curlew Bank fore reef, located just south of Carrie Bow Caye, is characterized by a low-relief spur-and-groove formation with colonies of the M. annularis complex, S. siderea, Diploria spp., and A. palmata (both living and dead colonies). Five surveys were in patch reefs. Mexico Rocks, seaward of Ambergris Caye, is an irregularly shaped complex of approximately 100 patch reefs at about 4 m depth. These patch reefs are dominated by large stands of the M. annularis complex (Burke et al., 1998), separated by sand and seagrass. Although this reef exhibited low coral diversity, live coral cover values were average for this reef type. Near the patch reef in the Hoi Chan Marine Reserve, which is dominated by Acropora palmata (both living and dead colonies) and the M. annularis complex, alO-m deep channel (Hoi Chan Channel) runs through the reef crest and into the back-reef region. The three patch reefs in the south-central region were at depths of about 2-5 m on the windward sides of Wee-Wee, Bread and Butter, and Norvall Cayes, where the M. annularis complex is the dominant coral species. Most colonies of A. palmata were “long dead” and were covered with turf algae. The three surveys in coral reef ridges were conducted at depths of 2-15 m off Wee-Wee Caye, Peter Douglas Caye, and an area locally known as “Tunicate Cove” within the Pelican Cayes. At these reefs most colonies of the dominant coral, A. tenuifolia, had died since the beginning of the 1998 warming event. Coral and algal populations were assayed by six-seven divers on each dive. Atlantic and Gulf Rapid Reef Assessment (AGRRA) Version 1 benthos protocols (see Appendix One, this volume) were used with the following modifications: stony corals >10 cm in diameter were included in the surveys; coral diameter and height were measured to the nearest cm for smaller corals (10-25 cm in diameter) and to the nearest 5 cm for larger (>25 cm) colonies. The Montastraea annularis complex was treated as a single species. Sediment deposits in the algal quadrats were removed by hand before estimating the abundance of crustose coralline algae. As space occupied by turf algae growing on live crustose corallines was allocated to both functional groups, total values for absolute abundance at some sites are in excess of 100%. Diadema antillarum, being 151 rare, was not counted. Training sessions were conducted with all divers [most of whom were already well-trained from the 1998 San Salvador Island, Bahamas AGRRA assessment (Peckol et al, this volume)] censusing “practice” transects on patch reefs. Species identifications, percent cover estimates, and coral disease and bleaching assessments were compared to ensure sampling consistency. We used Humann’s (1993) reef coral guide for coral species identifications; Littler and Littler (2000) was later consulted for macroalgae. A stationary visual census technique (Bohnsack and Bannerot 1986) was employed by two divers to survey the fish populations. All sampling occurred between 10:00 a.m. and 3:00 p.m. At each sampling point, all species belonging to eight families (Acanthuridae, Chaetodontidae, Haemulidae, Labridae, Lutjanidae, Pomacentridae, Scaridae, Serranidae) observed in five minutes within a 7.5 m radius cylinder were recorded. Each census was begun three minutes after laying a measuring tape on the substratum by counting all individuals of all species observed in the pre-set radius within the initial field of view. New sectors of fields of view were then scanned by rotating in one direction. Abundances of species moving in schools were taken when first observed in the sampling cylinder (it was important to count fishes moving in schools immediately because they were unlikely to remain in the sampling area). When very large schools were present it was sometimes necessary to estimate numbers in 10s or 50s. Fish lengths in cm were estimated using a T-shaped tool marked every 5 cm to help avoid underwater magnification problems (Bohnsack and Bannerot, 1986). We recorded the number of individuals, plus the minimum, maximum, and mean estimated lengths for each of the eight fish families. We used Humann’s (1994) reef fish guide for species identifications. Herbivorous fish grazing rates were measured by two divers at 1 2 sites (all but the Eagle Ray fore reef), following the AGRRA methodology given in Appendix One. Grazing rate measurements were all made between 10:00 a.m. and 2:00 p.m. during the peak time for grazing activity (Lewis, 1986). RESULTS Stony Corals Species composition. At each site we censused between 12-18 10-m transects, each of which generally had about 15 stony corals that were at least 10 cm in diameter (Table 1). At most sites we measured >250 coral colonies. Coral species abundance patterns varied greatly among the three reef types (Fig. 2A-C). Numerically, the most abundant coral species censused in the fore-reef spurs were Porites astreoides (22% of total) and the Montastraea annularis species complex (14% of total). Acropora palmata occurred here in relatively low abundance (<5% of total) but showed high total (recent and old) partial mortality of colony surfaces ranging from 25% (South Water Caye) to 100% (Tobacco Caye) and averaging >55%. Due to high wave-energy conditions, we were unable to sample the reef crest where A. palmata dominated (some colonies were alive). The M. annularis complex was clearly dominant in patch reefs (45% of total). Collectively, the mound corals, Siderastrea siderea, Diploria strigosa, and the M annularis complex represented ~60% of the corals sampled in patch reefs. Because their 152 colony sizes are generally larger, the numerically less abundant D. strigosa contributed more than 5'. siderea to reef habitat. Mostly dead Agaricia tenuifolia (30%) and P. astreoides (23%) dominated the south-central coral reef ridges. The M annularis complex was a minor component, and species richness was highest in this unique reef type (30 scleractinian species sampled along transects compared with ~24 species on patch reefs and fore reefs). A Fore Reefs, n = 1237 .■\Ci other 27% SS 10%. AGT IO“n DS 10® MA 14®o 0 PA 22®n AG —Agai 'ic ia agan cites .\QiT=Agciricia Tenuifolia DS=^Diploria sti igosa ]\l.\=AIonrastraea aiiiitilaris complex: VA.=Porites asti-eoides SS=Siclei astrea siderea I I Other @ T)^=DipIoria sti igosa |T~| W.\=Montastraea annularis complex m PP= Porites porites H PA=Porites astreoides SS=Siderastrea siderea I I Other B Patch Reefs, n Other lS®o SS 9®o PP(S®0 PA r«o = 1232 DS 5®o MA 4^% C Coral Reef Ridges, n = 723 A0"5®o PA 23“o (.'•ther 29“ 0 ON 7% MM "“o H □ AC=Acropora cerricornis AGT= Agaricia tenuifolia C\d=Colpopln Ilia n a tans MM=3 la dr acts iiiirah ilis PA=Porites asmeoides □ Other- Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>10 cm diameter) at (A) fore reefs; (B) patch reefs; (C) coral ridge reefs along the Belize barrier reef system. Other category = combined coral species, each with <5% abundance of occurrence. 153 Recruits. The most abundant of the coral recruits was Porites astreoides, which ranged from 31% of the total in the fore reefs to 50% in the patch reefs (Fig. 3A-C). Recruits of the M. annularis complex were rare (<2%) in the fore reefs and only represented 9% of the total in the patch reefs. Recruits of A. palmata were only found in A Fore Reefs, n = 1 14 B Patch Reefs, n = 146 Other 8% DS5% SS 6% FF5% PP 12% PA 50% C Coral Reef Ridges, n = 64 lill AG=Agaricia agairites ^ AGT=Agaricia tenuifolia Sgj DS=Diploria strigosa ^ PA=Poriles astreoides ■ ■ SSF=Sider€tstreasiderea m other □ m DS=Diploria strigpsa 'PY=Favia fragwn M.A=MotUastraea arwularis IV1C= Montastraea cavernosa PA= Porites astreoides □ VP=Porites poriles ^SF^iderastrea siderea Other AG'T=Agaricia tenuifolia MM=^Madracis mirabilis PA-Porites astreoides S\=Siephanocoenia intersepta Other Figure 3. Species composition and mean relative abundance of all stony coral recruits (<2 cm diameter) at (A) fore reefs; (B) patch reefs; (C) coral ridge reefs along the Belize barrier reef Other category = combined coral species, each with <5% abundance of occurrence. 154 the patch reefs but in extremely low abundance (<1% of total). A. tenuifolia was represented by 9% of the recruits in the ridges, but was more abundant (13% of total) in the fore reefs. Coral size. The mean sizes of individually surveyed corals (>10 cm in diameter) varied nearly threefold (22-60 cm) and were more variable among reefs in the fore reefs and reef ridges than in the patch reefs (Table 2). While most colonies of the M annularis complex in the fore reefs and patch reefs (and Siderastrea siderea in patch reefs) were <100 cm in diameter, we also measured a number of large, massive colonies (especially on the fore reefs), some being greater than 200 cm in diameter (Figs. 4A-B). A. palmata, although not abundant in the fore reefs, showed a large range in colony size from <20 cm to > 200 cm. A. tenuifolia in the reef ridges also showed a great range in size, but most colonies were <50 cm in diameter (Fig. 5A). cr u Urn A Fore Reefs Montastraea annularis complex (n=101) llll lllll I,,. .1, .1 00 Ov O — Size Distribution (cm) O' u B Patch Reefs Montastraea annularis complex (n=343) ll llilll-ii. .ll (N tn \o oo 10 cm diameter) at (A) fore reefs and (B) patch reefs along the Belize barrier reef. 155 A Size Size Distribution (cm) 180 160 140 120 100 80 60 40 20 0 B Partial Colony Mortality Percentage of Coral Colony Dead Figure 5. (A) Size-frequency distribution as % and (B) frequency distribution of total (recent + old) partial- colony mortality of Agaricia tenuifolia (>10 cm diameter) at coral reef ridges in south-central Belize. Coral condition. No diseased corals were found in the coral reef ridges and in the Hoi Chan patch reef. Elsewhere, signs of disease were relatively low (~2%) except in the South Water Caye fore reef where 5% of the censused corals were diseased (primarily black-band disease; Table 2). There was no clear difference in incidence of disease in fore reefs versus patch reefs but, considered together, the incidence of coral disease was twice as high in the south-central sites as in the northern region. Long-term effects from the 1998 warming event were still evident in May 1999. Bleaching affected >44% of the M annularis complex in the south-central patch reefs, and 42 % in both the northern and south-central fore reefs, while 26% of the colonies in the northern patch reefs were bleached. Fewer colonies (-15%) of each of the other common mound corals, Diploria spp. and Siderastrea siderea, were affected by bleaching. The relatively high percentages of bleached corals in most (4/5) of the fore reefs and all five patch reefs, particularly Norvall Caye, were obviously significant (Table 2). The percentages of bleached corals were higher on the south-central patch reefs (17- 40%) than in the corresponding northern reefs (15-16%). The fore reefs showed an opposite pattern with the northern reefs having proportionately more bleached corals than those in the south-central region (-22% versus -13%, respectively). Overall we found no clear spatial differences in extent of bleaching either between the fore reefs and patch reefs or between the northern and south-central regions (Table 2). Mortality characteristics of the corals censused at the 13 sites are summarized in Table 2. Recent partial-colony mortality ranged from -2% of upper surfaces in the fore reefs to sometimes >13% in the reef ridges. Comparable levels of old partial-colony mortality in the patch reefs and reef ridges were each about twice those in the fore reefs (- 35% versus -20%). Most colonies of Agaricia tenuifolia in the reef ridges were dead (Fig. 5B, with most of the A. tenuifolia in the 91-100% interval) and were covered by macroalgae and encrusting sponges (Chondrilla cf. nucula Schmidt; K. Riitzler, pers. comm.). While the percent of colonies scored as “standing dead” (100% mortality of upper colony surfaces and still in growth position) was 4% and 8% in the fore reefs and patch reefs, respectively, standing dead ranged from 18-50% in the coral reef ridges. 156 Total (recent + old) partial-colony mortality (hereafter total mortality) was considerably higher in the south-central ridges and patch reefs (both -45%) than in the northern patch reefs and south-central fore reefs (both -25%) and relatively low in the northern fore reefs (-15%; Table 2). At the regional scale, the north Belizean fore reefs and patch reefs had lower total mortality than similar reefs in the south-central region (-20% versus -35%, respectively). Total mortality for stony corals off Bread and Butter, Wee-Wee, and Norvall Cayes (mean = 45.5%, n = 3 reefs) was nearly two times greater than in the northern patch reefs. The greatest percentage was contributed by the M. annularis complex, for which total mortality values ranged between 34% (Bread and Butter) and >60% (Norvall, Wee-Wee) in the south-central patch reefs and from 35% (Hoi Chan) to 50% (Mexico Rocks) in the northern region. We plotted the frequency distributions of total mortality for sites located in northern (Fig. 6A,B) and south-central Belize (Fig. 7A-C) to illustrate these large-scale spatial differences. A majority of censused corals (>55-75% in patch reefs and fore reefs, respectively) in the northern region showed <10% total mortality of the colony surface. Although a substantial number of corals from the south-central region also had <10% total mortality, there was a greater proportion of colonies with higher percentages. For example, more than one-third of the censused corals at the coral reef ridges (n = 723) showed total mortality values of >90% of their surfaces (-75% of the colonies in this interval were A. tenuifolia). We also found a substantial number (>10%) of colonies with over 90% total mortality in the patch reefs (-28% of these colonies were the M annularis complex). A Northern Fore Reefs B Northern Patch Reefs Percentage of Coral Colony Dead Figure 6. Frequency distribution of total (recent + old) partial colony mortality of all stony corals (>10 cm diameter) at (A) fore reefs, (B) patch reefs in northern Belize. Number of Colonies Number of Colonies 157 A South-central Fore Reefs ON Percentage of Coral Colony Dead B South-central Patch Reefs ON Percentage of Coral Colony Dead C South-central Coral Reef Ridges Figure 7. Frequency distribution of total (recent + old) partial colony mortality of all stony corals (>10 cm diameter) at (A) fore reefs, (B) patch reefs, C) coral reef ridges in south- central Belize. 158 Algae The dominant component of the algal assemblage was turf algae in all patch reefs, all coral reef ridges, and one of the fore reefs. Macroalgae predominated in one fore reef, crustose corallines in a second, and the all three functional groups were approximately equally abundant in a third (Table 3). No difference was noted in macroalgal abundance between the northern and south-central fore reefs or patch reefs. Macroalgae in the fore reefs included the brown seaweeds, Dictyota humifusa, Lobophora variegata, and Stypopodium zonale, and the calcified green, Halimeda. Dominant macroalgae in the patch reefs were D. pulchella, Turbinaria turbinata, and the red alga, Galaxaura oblongata. Overall, mean macroalgal height was low (< 2 cm). Macroalgae indices (mean absolute macroalgal abundance x mean macroalgal height) were: <25 in two of the reef ridges, the two most southerly patch reefs and the most southerly fore reef; <100 at six sites; and >100 in one each of the south-central fore reefs and patch reefs. Fishes For the eight censused families, the highest fish abundances were in the Tunicate Cove ridge reef (mean = 105.1 individuals/200 m , se = 26). Mean fish densities in the five fore reefs ranged from 38.2 (se = 8.7) at San Pedro Canyon to 86 (se = 20.7) individuals/200m^ at Curlew but showed no regional trend (Table 1). The patch reefs showed relatively little spatial variation with fish densities ranging from 41.6 (se = 6.1) in Norvall to 60.3 (se = 12.2) individuals/200m^ in Bread and Butter. The two other reef ridges had low fish abundances (overall averaging about 30 individuals/200 m ). While damselfishes (Pomacentridae) were most abundant in the fore reefs (Fig. 8A), dominance was shared by four fish families in the patch reefs (Fig. 8B): surgeonfish (Acanthuridae), damselfish, parrotfish (Scaridae) and grunt (Haemulidae). Large (sometimes >100) schools of white grunt (Haemulon plumieri) associated with nearby seagrass habitat at Tunicate Cove ridge (Table 4) largely accounted for the high grunt densities in the reef ridges (Fig. 8C). Damselfish represented nearly a half (45%) of the total abundances in the other reef ridges. Parrotfish were more abundant than acanthurids in most (11/13) sites, and the latter were rare in the ridges. Only Mexico Rocks in northern Belize showed dominance by surgeonfishes over other herbivorous fishes. Densities of seabasses (Serranidae), snappers (Lutjanidae) and grunts varied greatly among sites (Table 4), ranging from combined totals of ~2 individuals/200 m (Wee-Wee reef ridge) to 86 individuals/200 m^ (Tunicate ridge). The latter also had the highest abundances of serranids and grunts, whereas snappers were most common in the Tobacco Caye fore reef Excluding Tunicate Cove, the Hoi Chan Marine Reserve patch reef showed highest combined densities (~30 individuals/200 m^) of these three families with snappers and grunts contributing the greatest proportion to that total. The size-frequency distributions for two major guilds of fishes, the herbivores (parrrotfishes, surgeonfishes, the yellowtail damselfish Microspathodon chrysurus) and carnivores (groupers and snappers) in each of the three reef types are shown in Figure 9. While the 6-10 cm length category clearly predominated in patch reefs and reef ridges for the herbivores, the 1 1-20 cm category was slightly more common in the fore reefs. Few herbivorous fishes were greater than 20 cm in length in any site. Although the majority 159 « 3 ]> O (U Si B 3 z 50 40 30 20 10 A Fore Reefs ■ , , -L , CD ' CO CO cO 2 2 2 S c 2 x: o g c CO O o CO o < cO x: u .i.ii U- r- ■S « ^ • rT> ►j 1/1 3 ’> 3 50 40 30 I 20 E Family B Patch Reefs C 03 'C s 03 J u 03 'S 03 •pi 3 ►J (U t) 03 03 "O "3 Family c e o cu CO o (72 Figure 8. Mean fish density (no. individuals/200 + se) for eight fish families at (A) fore reefs; (B) patch reefs; (C) coral ridge reefs along the Belize barrier reef. 160 1.00 "w 0.80 o Vm o a k. o a. o L. Cl. 0.60 0.40 0.20 0.00 1.00 I 0.80 o 0 0.60 U 1 0.40 V. CL, 0.20 0.00 0-5 A Fore Reefs ■ Herbivores n=78 ° Carnivores n=38 JLH 0-5 6-10 11-20 21-30 31-40 Length categories (cm) >40 B Patch Reefs ■ Herbivores n=80 '^Carnivores n=35 6-10 11-20 21-30 31-40 Length categories (cm) >40 C Coral Reef Ridges ■Herbivores n=41 Carnivores n=22 HL 6-10 11-20 21-30 31-40 Length categories (cm) >40 Figure 9. Size-frequency distribution of herbivores (all acanthurids and scarids, Microspathodon chrysurus) and carnivores (all lutjanids and serranids) at (A) fore reefs; (B) patch reefs; (C) coral ridge reefs along the Belize barrier reef 161 (80%) of the carnivores fell within the 6-10 cm or 1 1-20 cm length range, a greater range of sizes was present in the fore reefs and patch reefs than in the reef ridges. In particular, the patch reefs supported a substantial number (15% of the sample) of carnivores of the largest (>40 cm) size category, all of which were found within the Hoi Chan Marine Reserve. At the 12 sites for which we have measurements, herbivorous fish grazing rates ranged from 1.4 bites/min. (Peter Douglas reef ridge) to 13.4 bites/min. (Norvall Caye patch reef; Table 5). Overall, grazing rates were slightly higher in the patch reefs (mean = 6.4 bites/min., n = 5 reefs) than in the fore reefs (mean = 3.2 bites/min., n = 5 reefs). Generally, parrotfishes and surgeonfishes were the predominant grazers in patch reefs and fore-reefs, respectively. No relationship was found between herbivorous fish density and macroalgal index (Fig. 10) nor between grazing rate and macroalgal index (regression analysis, F<1). 140 120 I 100 Sf 80 ^ 60 ■ ® 40 ■ s 20 ■ 0 ^ ^ ^ ^ ^ ^ ' 0 10 20 30 40 50 60 70 Mean Herbivore Density (individuais/200m^) Figure 10. Regression plot between mean herbivore density (no. individuals/200 m^) and mean macroalgal index, by site along the Belize barrier reef. DISCUSSION Our May, 1999 AGRRA surveys along the northern and south-central regions of the Belize barrier reef were conducted approximately seven-eight months after the occurrence of two large-scale disturbances, a major thermal anomaly and the passage of Hurricane Mitch. Many corals had not yet recovered from the September 1998 warming event. Although the fore reefs and patch reefs appeared to have lower recovery rates than the coral reef ridges, this is only because Agaricia tenuifolia, the species most affected by the thermal anomaly, had sustained high rates of total colony mortality (see also Aronson et ah, 2000). In the south-central patch reefs, 50% of the colonies of another major reef 162 builder, the Montastraea annularis complex had been bleached the previous January (Peckol et ah, 2001). Over 44% of this population was still affected in May, 1999 when comparable percentages in the northern patch reef averaged 26%. However, earlier bleaching data are lacking from these northern sites, and we do not know if this regional difference represents more rapid recovery or is related to a lower initial stress during the thermal anomaly. Other common mound corals {Diploria spp. and Siderastrea siderea) showed higher incidences of recovery; whereas ~34% had been bleached in January, only 1 5% of all species remained discolored in May in south-central sites. In May 1 999, 42% of the M annularis complex in our fore reefs had discolored tissues. The following month Kramer and Kramer (2000) found significant remnant bleaching (to >40% of the colonies) at depths of 8-20m in 36 fore reefs off Belize. In contrast, Mumby (1999) reported that though 70-90% of coral colonies in the fore reef (at 8-10 m) of Glovers Atoll (Belize) appeared to be either fully or partially bleached in response to the 1998 warming event, these colonies rapidly (by late December 1998) regained their usual coloration . A number of studies have reported that recovery from bleaching proceeds more slowly in corals occurring at greater depths (see Kramer and Kramer, 2000). The long- term impacts and potential for recovery of the Belize barrier reef from such a large-scale thermal anomaly are uncertain. Generally, our survey of the Belize barrier reef suggests a degradation of this once largely pristine reef environment (Perkins and Carr, 1985; Cortes, 1997). Such deterioration may stem from a combination of direct and indirect effects from global climate change (Aronson et al., 2000) and regional human activities (Higinio and Munt, 1993; Carter et al., 1994). Within-region partial mortality rates of stony corals were higher in lagoonal reefs than in corresponding fore reefs, suggestive of greater effects overall from the 1998 coral bleaching event than from Hurricane Mitch. The greatest impacts from this major storm were found in fore reefs and outer atoll reefs off Belize (Mumby, 1999). Kramer and Kramer (2000) reported major damage at several sites off Belize where damage from Hurricane Mitch resulted in a notable loss of reef structure; they noted storm damage down to their deepest survey site at 20 m in the fore reef. During a previous AGRRA survey in January 1999 in a fore reef near South Water Caye (Peckol et al., unpublished data), we documented breakage of branching corals, including Millepora alcicornis (33%), M complanata (10%), and Porites porites (30%). We also recorded toppling of about 3% of the colonies of the M annularis complex. Substantial amounts of rubble and entire mound corals had been transported onto and behind the reef crest. By May 1999, however, our AGRRA survey found limited evidence of the hurricane effects on the attached corals occurring in fore reefs. Because major frame-builders, such as the Montastraea annularis complex and Acropora palmata, were rare as recruits in our sites, reef recovery may be slow following major disturbances like hurricanes and widespread bleaching. For example, although the M. annularis complex represented nearly 50% of the colonies >10 cm diameter in the patch reefs, it showed relatively low (9%) recruit abundance. Moreover, increasing macroalgal and sponge cover on the dead colonies of Agaricia tenuifolia in the coral reef ridges following the thermal anomaly (Peckol et al., unpublished data) may limit coral recruitment onto this substrate. Aronson and Precht (2001) reported that Echinometra viridis had suppressed macroalgal growth after the mass mortality of A. cervicornis from 163 white-band disease in the 1980s, allowing recruitment of^. tenuifolia. Although Porites astreoides was the dominant recruit (nearly 50% numerically) in this habitat, we did find some recruits (9% of total) of the formerly dominant A. tenuifolia. Turf algae dominated in patch and coral reef ridges, yet macroalgae were quite prevalent representing >30% absolute abundance in the quadrats at 6/13 sites and exceeding 40% in two of the fore reefs (Table 3). Many Caribbean reefs have experienced exponential increases in macroalgal cover (Hughes, 1994; Shulman and Robertson, 1996; Peckol et ah, this volume) often attributed to reductions in the sea urchin, Diadema antillarum, and herbivorous fishes (Hughes, 1994; Hughes et ah, 1999). However, densities of D. antillarum were relatively low in the Belize barrier reef before the mass mortality event (Hay, 1984) and, until recently, much of this reef system has experienced low fishing pressures. Rapidly increasing abundances of brown macroalgae have been reported in a number of fore reefs in several studies of the Belize barrier reef (Littler et ah, 1987; Aronson et ah, 1994), including areas considered relatively remote from human activities (McClanahan et ah, 1999). The local cause(s) of such change for the Belize barrier reef remains unclear although high rates of coral mortality, the two warming events of the 1990s, nutrient loading, and reduced herbivory from fishes have been suggested (Aronson and Precht 1997; McClanahan et ah, 1999; Aronson et ah, 2000). Because many areas experiencing increased macroalgal cover are remote from human activities, nutrient loading is unlikely to be the direct cause of such changes (McClanahan et ah, 1999). The grazing rates measured in the present study were approximately twice as high in patch reefs as in fore reefs. In addition, mean parrotfish densities were somewhat greater in south-central patch reefs compared with other sites in Belize and grazing by parrotfish represented >50% of the observed activities in these patch reefs where turf algae still dominate. Parrotfish densities overall (-5.2/200 m^ and in leeward patch reefs (~6.5/200m^) documented in a 1998 AGRRA survey off San Salvador Island, Bahamas (Peckol et ah, this volume) were about half those measured in the Belize barrier reef (overall 1-10.9/200 m ; patch reefs -1 1.2/200 m ). Absolute macroalgal abundances were approximately two (44% overall) to three (59% on patch reefs) times greater off San Salvador Island than in Belize (-25%). Similar to Belize, the San Salvador reefs are also experiencing increases in cover by brown seaweeds. Lewis (1985) reported that parrotfishes actively grazed on several brown algal genera not consumed by surgeonfishes. Perhaps changing abundance patterns of parrotfishes, particularly in fore reefs, in part has contributed to increases in brown macroalgae in the Belize barrier reef Further study should allow us to document such a relationship. The Hoi Chan patch reef had the highest combined densities of carnivorous fishes (seabasses, snappers, grunts) of all the surveyed patch reefs and the largest carnivores were found within the reserve. Partial-coral mortality was lower in the northern region than at the south-central sites. These spatial differences in reef condition (coral colony mortality, fish population abundance, and size) in part may be related to the establishment of the Hoi Chan Marine Reserve in 1987. Prior to its establishment, the area was subject to uncontrolled fishing pressures that had removed most of the large and mobile fish from the reef and to burgeoning numbers of tourists who often damaged its corals. With the establishment of the reserve, these issues were largely addressed (e.g., installation of mooring buoys, education of tourists by local guides about behavior while 164 diving on reefs), and the community gained a critical conservation ethic and pride in their unique marine resource (Carter et ah, 1994). Recent studies, which included the Hoi Chan Marine Reserve (Sedberry et ah, 1992; Roberts, 1995), have demonstrated that snapper and grouper populations are more abundant and fishes are larger within reserves than outside their boundaries. The added benefit of marine reserves is that they can result in increased fish abundances in adjacent reefs as a “spill-over effecf ’ (Russ and Alcala, 1996). Clearly, this small but successful reserve admirably serves as a model for future reef conservation for the Belize barrier reef system as well as elsewhere. ACKNOWLEDGMENTS This research was funded by a Culpeper Foundation grant to Smith College (Peckol and Curran, co-principal investigators), by the Smith College Summer Science Program, and by the B. Elizabeth Homer Fund. We greatly appreciate the field- and data- processing assistance provided by Smith College AGRRA team members: Elise Cormier, Carrie Fraga, Amelia Jugovich, Rachel Luksic, and Shannon Ristau. We thank Wee Wee Caye Marine Laboratory for facilities and staff support. REFERENCES Aronson, R.B., and W.F. Precht 1997. Stasis, biological disturbance and community structure of a Holocene coral reef Paleobiology 23:326-346. Aronson, R.B., and W.F. Precht 2001. Evolutionary paleoecology of Caribbean coral reefs. Pp. 171-233. In: W.D. Allmon and D.J. Bottjer (eds.). Evolutionary Paleoecology: The Ecological Context of Macroevolutionary Change, Columbia University Press, New York. Aronson R.B., P.J. Edmunds, W.F. Precht, D.W. Swanson, and D.R. Levitan 1994. Large-scale, long-term monitoring of Caribbean coral reefs: simple, quick, inexpensive techniques. Atoll Research Bulletin 42:1-19. Aronson R.B., W.F. Precht, LG. Macintyre, and J.T. Murdoch 2000. Coral bleach-out in Belize. Nature 405:36. Bohnsack, J.A., and S.P. Bannerot 1986. A stationary visual census technique for quantitatively assessing community structure of coral reef fishes. NOAA Technical Report National Fish and Wildlife Service 41:1-15. Burke C.D., W.D. Bischoff, S.J. Mazzullo, and T.M. McHenry 1996. Coral mortality associated with the 1995 western Caribbean bleaching event, Mexico Rocks patch reef complex, Belize. Geological Society of America, annual meeting 28:274 (Abstract). Burke C.D., T.M. McHenry, W.D. Bischoff, and S.J. Mazzullo 1998. Coral diversity and mode of growth of lateral expansion patch reefs at Mexico Rocks, northern Belize shelf. Central America. Carbonates and Evaporites 13:32-42. 165 Carter J., J. Gibson, A. Carr III, and J. Azueta 1994. Creation of the Hoi Chan Marine Reserve in Belize: a grass-roots approach to barrier reef conservation. Environmental Professional 16:220-231. Cortes J. 1997. Status of the Caribbean coral reefs of Central America. Proceedings of the Eighth International Coral Reef Symposium, Panama City, 1:335-340. Ginsburg, R.N. (compiler) 1 994. Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards, and History, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 420 pp. Hay M.E. 1984. Patterns of fish and urchin grazing on Caribbean coral reefs: are previous results typical? Ecology 65:446-454. Higinio E., and I. Munt 1993. Belize: eco-tourism gone awry. Report on the Americas. 26:8-10. Hughes, T.P. 1994. Catastrophes, phase shifts and large-scale degradation of a Caribbean coral reef Science 265:1547-1550. Hughes, T.P., A.M. Szmant, R. Steneck, R. Carpenter, and S. Miller 1999. Algal blooms on coral reefs: what are the causes? Limnology and Oceanography 44:1583-1586. Humann, P. 1993. Reef Coral Identification. New World Publications, Inc., Jacksonville, FL, 252 pp. Humann, P. 1 994. Reef Fish Identification, 2"^ ed. New World Publications, Inc., Jacksonville, FL,'424 pp. James N.P., and R N. Ginsburg 1 979. The Seaward Margin of Belize Barrier and Atoll Reefs: Morphology, Sedimentology, Organism Distribution and Late Quarternary History. International Association of Sedimentologists, Special Publication No. 3. 196 pp. Kramer, P.A., and P.R. Kramer 2000. Ecological Status of the Mesoamerican Barrier Reef System: Impacts of Hurricane Mitch and 1998 Coral Bleaching. Final Report to the World Bank, 73 pp. Kramer, P.A., P.R. Kramer, A. Arias-Gonzalez, and M. McField 2000. Status of coral reefs of northern Central America: Mexico, Belize, Guatamala, Honduras, Nicaragua and El Salvador. Pp: 287-313. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000, Australian Institute of Marine Science. Cape Ferguson, Queensland and Dampier, Western Australia. Lapointe, B.E. 1997. Nutrient thresholds for bottom-up control of macroalgal blooms on coral reefs in Jamaica and southeast Florida. Limnology and Oceanography 42:1119-1131. Lapointe, B.E. 1999. Simultaneous top-down and bottom-up forces control macroalgal blooms on coral reefs (reply to the comment by Hughes et al.). Limnology and Oceanography 44:1586-1592. 166 Lessios, H.A. 1988. Mass mortality of Diadema antillarum in the Caribbean: what have we learned? Annual Review of Ecology and Systematics 19:371-393. Lewis, S.M. 1985. Herbivory on coral reefs: algal susceptibility to herbivorous fishes. Oecologia 65:70-375. Lewis, S.M. 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Littler, D.S., and M.M. Littler 2000. Caribbean Reef Plants. An Identification Guide to the Reef Plants of the Caribbean, Bahamas, Florida and Gulf of Mexico. Offshore Graphics, Inc., Washington, D.C. 542 pp. Littler, M.M., P.R. Taylor, D.S. Littler, R.H. Sims, and J.N. Norris 1987. Dominant macrophyte standing stocks, productivity and community structure on a Belizean barrier reef Atoll Research Bulletin 302:1-24. MacIntyre LG., and R.B. Aronson 1997. Field guidebook to the reefs of Belize. Proceedings of the Eighth International Coral Reef Symposium, Panama City 1 :203-222. Macintyre, LG., W.F. Precht, and R.B. Aronson 2000. Origin of the Pelican Cays Ponds, Belize. Atoll Research Bulletin 466:1-1 1. McClanahan, T.R., R.B. Aronson, W.F. Precht, and N.A. Muthiga 1999. Fleshy algae dominate remote coral reefs of Belize. Coral Reefs 18:61-62. McField, M.D. 1999. Coral response during and after mass bleaching in Belize. Bulletin of Marine Science 64:155-172. McGrath, T.A., and G.W. Smith 1999. Monitoring the 1995/1996 and 1998/1999 bleaching events on patch reefs around San Salvador Island, Bahamas. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration, Fort Lauderdale, p. 135-136. Mumby, P.J. 1999. Bleaching and hurricane disturbances to populations of coral recruits in Belize. Marine Ecology Progress Series 190:27-35. Nemecek, S. 1999. Whiteout: widespread coral bleaching, even in deep waters, continues to perplex scientists. Scientific American April:30-3L Peckol, P., H.A. Curran, M. Robbart, and B. Greenstein 2001 . Resilience and recovery of coral reefs from large-scale disturbances: contrasting patterns for San Salvador Island, Bahamas and Belize. Pp: 133-145. In: B. Greenstein. and C. Carney (eds.). Proceedings of the 10‘^ Symposium on the Geology of the Bahamas and other Carbonate Regions. Bahamian Field Station, San Salvador, Bahamas. Perkins, J.S., and A. Carr III 1985. The Belize barrier reef: status and prospects for conservation management. Biological Conservation 3 1 :29 1-301. 167 Roberts, C.M. 1995. Rapid build-up of fish biomass in a Caribbean marine reserve. Conservation Biology 9:815-826. Russ, G.R., and A.C. Alcala 1996. Do marine reserves export fish biomass; evidence from Apo Island, central Philippines. Marine Ecology Progress Series 132:1-9. Riitzler K., and I.G. MacIntyre, (editors) 1982. The Atlantic Barrier Reef Ecosystem at Carrie Bow Cay, Belize, 1. Structure and Communities. Smithsonian Contributions to Marine Sciences, Smithsonian Institution Press, Washington DC, No 12, 539 pp. Sedberry, G., J. Carter, and P. Barrick 1992. The effects of fishing and protective management on coral reefs of Belize. Proceedings of the Gulf and Caribbean Fisheries Institute 1 1 : 1-25. Shulman M.J., and D.R. Robertson 1996. Changes in the coral reef of San Bias, Caribbean Panama: 1983 to 1990. Coral Reefs 15:231-236. Smith, S.V., and R.W. Buddemeir 1992. Global change and coral reef ecosystems. Annual Review of Ecology and Systematics 23:89-1 18. Stoddart, D.R. 1974. Post-hurricane changes on the British Honduras reefs: re-survey of 1972. Proceedings of the Second International Coral Reef Symposium 2:473-483. Wilkinson, C. 2000. Executive summary. Pp. 7-17. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science, Cape Ferguson, Queensland and Dampier, Western Australia. Table 1. Site information for AGRRA stony coral, algal and fish surveys off Belize, May 1999. 168 169 (U _N 13 PQ M— I o (U X) 25 cm in diameter) stony corals, the density of stony coral recruits, and the scarcity of benthic macroalgae. However, concerns were raised about the fringing reef, particularly off the island’s southern coast. Although the causes for its poor condition here are not well understood, the intrinsic oceanographic setting, in particular exposure to storm waves in winter, and the presence of numerous tourist divers and snorkelers during the summer must be investigated. INTRODUCTION The southernmost coral reefs in the western Atlantic Ocean are found in Brazil (Fig. 1). They are spread over a distance of approximately 2,000 km between 0°50' and 19° S latitude. Although some information about the Brazilian reefs has existed for over a century, many reef areas are still poorly known and there are few quantitative assessments of their condition. Along the coast of the state of Bahia (between 12 °30' and 1 9° S latitude), where most coral reefs occur, few previous reef assessments have been performed (Leao et al., 1997; Castro and Pires, 1999; Leao et al., 1999). In the southern part of Bahia state (at 17°-19° S latitude), the continental shelf widens to 200 km forming the Abrolhos Bank, where the largest and richest of Brazil’s coral reefs are scattered over ’ Former address: Universidade Estadual de Feira de Santana, BRl 16, Km3, Campus Universitario, Feira de Santana 44031-460, Bahia, Brazil. 'S Universidade Federal da Bahia, Laboratorio de Estudos Costeiros, Rua Caetano Moura 123, Salvador 40210-340, Bahia, Brazil. Email: rkikuchi@cpgg.ufba.br Former address: Parque Nacional Marinho de Abrolhos, IBAMA, Praia do Kitongo s/n, Caravelas, Bahia, Brazil. 174 an area of approximately 6,000 km (Fig. 1). Isolated bank reefs of varied shapes (elongated, circular, etc.), with dimensions that range from less than 1 km to about 20 km in length, occur nearshore in depths of 10-20 m. Shallow offshore (<10 m deep) fringing reefs border the islands of the Abrolhos Archipelago, and enormous (maximally to 25 m high) isolated mushroom-shaped reef columns, or “chapeiroes” (Fig. 2), are found in deeper water. The morphology of these major reef types seems to be related to the underlying substratum (older reefs, Precambrian bedrock, beachrock, etc.), the prevailing hydraulic regime, and their position relative to present sea level (Leao, 1996; Kikuchi and Leao, 1998; Leao and Kikuchi, 1999; Leao et al., 2001). Figure 2. Cross-sectional sketch of the mushroom-like growth form of the Brazilian chapeiroes. The Abrolhos reefs are built by a stony coral fauna that is characterized by its very low diversity compared with those of the North Atlantic or Indo-Pacific Oceans, by the endemic character of its major reef-builders, and by the complete absence of species with branching morphologies (Belem et al., 1986; Castro, 1994). Of the 18 species identified so far in Brazilian reefs, 17 are reported from the Abrolhos region. Most common are the six endemic scleractinians (Mussismilia braziliensis, M. hispida, M. harttii, Siderastrea stellata, Favia gravida and F. leptophylla), some of which have affinities with modem Caribbean species, while others are related to a Tertiary coral fauna that may have been isolated from the Caribbean Sea by the Amazon River flow, after its reversal due to the elevation of the Andes during the Tertiary. The hydrocoral Millepora alcicornis occupies reef-edge habitats, but all Brazilian reefs lack the acroporids which are major components of many reef-crest and fore-reef zones in North Atlantic reefs. Stony corals inhabiting the nearshore bank reefs are naturally exposed to influxes of terrigenous sediment (Martin et al., 1985). They are now threatened by accelerated urban expansion which has caused an increase in coastal mnoff and untreated sewage discharges. The offshore reefs, in contrast, are located away from the mainland and are partially protected by law (see Kikuchi et al., this volume). Previous assessments of potential human impacts in the Abrolhos National Marine Park have referred only to the 175 anchorage of tourist boats in seagrass beds around the islands (Creed and Amado Filho, 1999). The northern coasts of the Abrolhos islands (Fig. 1) are exposed to relatively high wave energies between September and February, their southern sides to somewhat larger waves from April to August, while the western (leeward) coasts are protected from major wind trends. Their fringing reef flats extend about 30 m from shore, are poorly developed and, during low tides, are subaerially exposed. Seaward, the reefs gradually slope to their edges at depths of 4-5 m where they drop off to a sandy bottom at 8-10 m. The shallow fringing reefs around some of the islands are intensively used for recreational scuba diving and snorkeling. In particular, tourists dive all around Santa Barbara Island (SAB), most frequently on its southern side. The windward Parcel dos Abrolhos (PAB) chapeiroes, which are located about 2 km east of the Abrolhos islands, form huge mushroom-shaped structures, some of which attain heights of over 20 m. Due to the limited penetration of light beneath the overhangs around their sides, the maximum development of reef corals is at depths of 5-8 m on the tops of the chapeiroes. The PAB chapeirdes, which are seldom visited by tourists due to difficulty of navigation and rougher sea conditions, are known to be in good condition (personal observations). Our initial purpose in using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocols in the Abrolhos National Marine Park was to collect information on the condition of its offshore reef communities in order to establish standards for a long- term monitoring program in the park. The fish community assessment results can be found in Kikuchi et al. (this volume). METHODS Benthic surveys were conducted in eight fringing reef sites around SAB that are considered representative of its northern (three sites), southern (three sites) and western (two sites) coasts. Each site was located in shallow water (3. 5-5. 5 m) as close as possible to the seaward edge of the reef The three sites off the southern coast were also strategic choices because they are preferred recreational dive sites. The crests of five representative chapeiroes, located at depths of 6-6. 5m, were surveyed. All of these sites are routinely exposed to large waves, except for the leeward (western) fringing reef, which experiences strong tidal currents. Two divers utilized the AGRRA Version 2.0 benthos protocol (see Appendix One, this volume) on each dive. Damselfish algal gardens were not assessed; rather the fish assessment team recorded the sizes and densities of all damselfishes. The average maximum height of all the “large” (>25cm in diameter) corals surveyed in each transect was used to approximate the rugosity of the reef surface. Two days were spent in consistency training before the beginning of the surveys. Reef environments (the windward fringing reef off southern and northern SAB, the leeward fringing reef off western SAB, and the chapeiroes) were compared on the basis of average live stony coral cover, number of large corals per transect, average maximum height of large corals, and number of large corals affected by old partial- colony mortality. Although tests of normality showed that the data did not depart from 176 normal distributions, most of the variances are rather unequal. Thus we chose to perform a Kruskal-Wallis non-parametric ANOVA for multisample hypothesis testing (Zar, 1999). When the null hypothesis was rejected by the Kruskal-Wallis test, we applied a non-parametric Tukey-type method to locate the differences among the data, using the test proposed by Dunn (see Zar, 1999). RESULTS Stony Corals Live stony coral cover averaged about 20% in the chapeiroes (Table 1) and was significantly higher than in the SAB fringing reef (Kruskal-Wallis, < 0.001), where corresponding cover estimates ranged from 3.5-12%. The density of large (>25 cm in diameter) stony corals varied between ~0.4/m (southern SAB) and ~l.l/m (western SAB) and was significantly smaller in the southern SAB sites (Kruskal-Wallis,/? < 0.001). Less variation was seen in the maximum height of the large corals with the tallest, averaging ~30 cm, occurring in the chapeiroes (Table 2). However, stony corals in the northern SAB sites and the chapeiroes are significantly taller than those in the southern and western SAB sites (Kruskal-Wallis,/? < 0.001). The mean diameter of the large corals was quite similar in the four surveyed environments varying from about 37 to 48 cm (Table 2). Six species accounted for more than 95% of the large corals. In all sites, Mussismilia braziliensis was by far the dominant species (comprising >66% of all colonies) especially in the western and northern SAB sites (Table 3, Fig 3). Millepora alcicornis and Montastraea cavernosa were present in all sites but the former was more common in the chapeiroes and the latter in the fringing reef Mussismilia hartti was slightly more abundant in the chapeiroes than in the fringing reef Favia leptophylla was found in all sites, while Siderastrea stellata was absent from the western side of the island. Size distributions of Mussismilia braziliensis, shown in Figure 4, were unimodal in the western and northern SAB fringing reef and the chapeiroes, with the modes in the 30-40 cm interval. However, colonies were more evenly distributed across size classes in the southern SAB sites. Corals greater than 70 cm in diameter were more frequent in the fringing reef (where they formed more than 1 0% of the population) than in the chapeiroes (Fig. 4). Recent partial-colony mortality (hereafter recent mortality) of large stony corals ranged from 0 to 10% with the highest overall averages (-5-6.5%) in the chapeiroes (Table 2, Fig. 5). Estimates of recent mortality for Mussismilia braziliensis resembled those of the entire assemblage of large corals except for one southern SAB site and one of the chapeiroes where no colonies of this dominant species had incurred any recent tissue loss (Table 3). Mean estimates of old partial-colony mortality (hereafter old mortality) for large corals overlapped in the four surveyed environments (individual site averages varied twofold, between -17 and 32%) as shown in Figure 5, although proportionately fewer 177 Figure 3. Species composition and mean relative abundance of all stony corals (>25 cm diameter) in the (A) southern and (B) western and northern SAB fringing reef, (C) PAB chapeiroes, in the Abrolhos, Brazil. FL = Favia leptophyla, MC = Montastraea cavernosa, MLA = Millepora alcicornis, MB = Mussismilia braziliensis, MHA = M hartti, SST = Siderastrea stellata. 178 30-39 50-59 70-79 90-99 110-119 Maximum diameter interval (cm) Maximum diameter interval (cm) Maximum diameter interval (cm) Figure 4. Size-frequency distribution of colonies (>25 cm diameter) of Mussismilia braziliensis in the (A) southern and (B) western and northern SAB fringing reef, and (C) PAB chapeiroes in the Abrolhos, Brazil. 179 Percent surface affected Western and northern SAB fringing reef Percent surface affected 100 CO 0 'c o o o c 0 o 0 Q_ PAB chapeiroes 0-10 20-30 40-50 60-70 80-90 10-20 30-40 50-60 70-80 90-100 Percent surface affected Figure 5. Frequency distribution of recent partial colony mortality and old partial colony mortality of all stony corals (>25 cm diameter) in the (A) southern, (B) western and northern SAB fringing reef, and (C) PAB chapeiroes in the Abrolhos, Brazil. 180 colonies were affected in the chapeiroes than in the fringing reef (-35-62% versus -64-86%, respectively) (Table 2). The number of colonies with old mortality was significantly greater in the western SAB sites than in the southern SAB sites and the chapeiroes (Kruskal- Wallis,/? < 0.001). Mean estimates of total (recent + old) partial-colony mortality (hereafter total mortality) exhibited similar patterns to those described for old mortality (Table 2) by overlapping in the four surveyed environments with proportionately fewer colonies being affected in the chapeiroes than in the fringing reef (-38-64% compared to 64-84%, respectively). Similar patterns for old mortality and total mortality characteristies were seen in Mussismilia braziliensis (Table 3). No signs of disease were seen in any of the large stony corals in the survey sites (Table 2). No bleached colonies were observed in the northern SAB sites and they were rare (0-3.5% of corals) off the southern and western sides of the island. Levels of bleaching were slightly higher in the chapeiroes (1.5-7%). The density of coral recruits was higher in the chapeiroes than in the fringing reef (Table 4). The most common species found as recruits in both habitats were Siderastrea stellata, Agaricia agaricites, Favia gravida and, very rarely, Mussismilia braziliensis. Algae and Diadema antillarum Turf algae clearly predominated everywhere except in one of the northern SAB sites (SAB8), and were particularly abundant (>80% relative abundance) in the other two northern sites (Table 4). The relative abundance of macroalgae was higher overall in the fringing reef (<0.5-58%, n=8 sites), where they formed the dominant algal group in one northern site (SAB8), than in the chapeiroes (<0.5-7%, n=5 sites). The average height of the macroalgae and the maeroalgal index (relative macroalgal abundance x macroalgal height) were also greater in all but one of the SAB sites than in the ehapeiroes. Crustose coralline algae were relatively more abundant in the chapeiroes than in the fringing reef. No individuals of Diadema antillarum were found during the surveys. DISCUSSION Although the erests of the chapeiroes were slightly deeper (6-6.5 m) than the fringing reef (3. 5-5. 5 m), most sites were similar in terms of the diameter and old partial- mortality of large (>25 cm) stony corals and in the predominance of turf algae on the substratum. Nevertheless, considering their higher values for live stony coral cover, maximum coral height and coral recruit density, as well as their low maeroalgal indices (a proxy for macroalgal biomass), the chapeiroes appeared to be in better condition overall than the SAB fringing reef. The worst indicators were found on the southern fringing reef where the density of large corals and live stony coral cover are both particularly low. It should be remembered that the southern SAB reef is most heavily visited by tourists (divers and snorkelers) during the summer. Species richness among large stony corals may be higher in the more exposed southern fringing reef and the chapeiroes than in the western and northern SAB sites. Storms may prevent Mussismilia braziliensis (which, due to its characteristic mushroom 181 shape, breaks and tumbles easily) from dominating other species. Although large (>40 cm) colonies of M braziliensis are relatively more common in the southern SAB sites where coral density is low, we found no colonies greater than 1 10 cm in diameter in this environment. Intense fragmentation and subsequent mortality of the fragments may reduce the abundance both of the small and of the very large (>110 cm) size classes; alternatively recruitment and/or survival of young colonies may be less successful in this habitat. So far neither outbreaks of disease nor any mass mortality events have been witnessed among the Brazilian stony coral populations. The somewhat higher values of recent partial-colony mortality and the slightly higher percentage of bleached corals that were present in the chapeiroes presently lack an explanation. On the other hand, the percentage of colonies with old partial-mortality was substantially higher in the fringing reef than in the chapeiroes (Table 2). That the southern SAB sites were not significantly different from the chapeiroes can be tentatively explained by the scarcity of large corals (i.e., small sample sizes) off this side of the island. The density of coral recruits was inversely related to the relative abundance of macroalgae, hence higher in the chapeiroes than in the SAB fringing reef. The dominance that we found by recruits of Siderastrea and Agaricia is expected since the life history strategies of many species in these genera are to concentrate energy in reproduction rather than long-term survival (Bak and Engel, 1979). Clearly, this is not the case for M. braziliensis', nonetheless, a few of its recruits were present in this survey. The higher relative abundance of macroalgae and the much higher macroalgal index values found in the SAB fringing reef compared to the more isolated chapeiroes are also expected, given their proximity to the island. In particular, sewage seep from the houses (near SABS) and imputs from the nesting birds and/or the decomposed remains of dead birds in the “bird’s cemetary” on the northwestern coast (near SABS), would augment nutrients in natural runoff during storms (see Kikucki et al., this volune). Grazing pressures could also be higher in the chapeiroes: Echinometra lucunter, which is known to be a significant herbivore, is more abundant than Diadema antillarum along the eastern coast of Brazil, but its impact in the offshore Abrolhos reefs is not known. However, E. lucunter is more common in the fringing reef than the chapeiroes (personal observations), as are key herbivorous surgeonfishes and parrotfishes although the latter are smaller off SAB than in the PAB (Kikuchi et al., this volume). An overall impression given by the AGRRA benthos indicators is that reef condition is fairly good in the chapeiroes and in the northern SAB fringing reef, although the cover of live stony corals is low in the latter. Interestingly, the two reef areas that appeared to be in poor condition (i.e., the southern and, to a lesser extent, the western SAB coasts) include the sites preferred by tourists for diving and snorkeling. Thus, closer monitoring and evaluation of the relative impacts of summertime tourism and winter storm damage are warranted. To have a more complete picture of the entire Abrolhos region, the offshore islands that are inaccessible to visitors will be assessed for comparison with the recreational dive sites at SAB. We will investigate the condition of some chapeiroes located near the archipelago which are popular with tourist divers and snorkelers. We also plan to apply the AGRRA protocols in the coastal reefs that have been most severely impacted by anthropogenic effects. 182 ACKNOWLEDGMENTS The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. The authors acknowledge the logistical support offered by the Abrolhos National Marine Park during the field work and the financial support of the Geology Graduate course of Federal University of Bahia headed by Dr. Geraldo S. Vilas Boas. The AGRRA editorial committee offered important suggestions to the final version of this manuscript. REFERENCES Bak, R.P.M., and M.S. Engel 1979. Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Marine Biology 54:341-352. Belem M.J.C., C. Rohlfs, D.O. Pires, and C.B. Castro 1986. S.O.S. Corais. Ciencia Hoje 5:34-42. Castro C.B. 1994. Corals of Southern Bahia Pp. 161-176. In: B. Hetzel and C.B. Castro (eds.), Corals of Southern Bahia. Editora Nova Fronteira, Rio de Janeiro. Castro C.B., and D.O. Pires 1999. A bleaching event on a Brazilian coral reef. Revista Brasileira de Oceanografia, 47:87-90. Creed J. C., and G.M. Amado Filho 1999. Disturbance and recovery of the macroflora of a seagrass (Halodule wrightii Ascherson) meadow in the Abrolhos Marine National Park, Brazil: an experimental evaluation of anchor damage. Journal of Experimental Marine Biology and Ecology 235:285-306. Kikuchi R.K.P., and Z.M.A.N. Leao 1998. The effects of Holocene sea level fluctuation on reef development and coral community structure. Northern Bahia, Brazil. Anais da Acaemia Brasileira de Ciencias 70:159-171. Leao Z.M.A.N. 1996. The coral reefs of Bahia: morphology, distribution and the major environmental impacts. Anais da Acaemia Brasileira de Ciencias 68:439-452. Leao Z M.A.N., and R.K.P. Kikuchi 1999. The Bahian coral reefs - from 7000 years BP to 2000 years AD. Ciencia e Cultura - Journal of Brazilian Association for the Advancement of Science 51: 62-273. Leao Z.M.A.N., R.K.P. Kikuchi, M.P. Maia, and R.A.L. Lago 1997. A catastrophic coral cover decline since 3,000 years B.P., Northern Bahia, Brazil. Proceedings of the Eighth International Coral Reef Symposium 1:583-588. 183 Leao, Z,, R. Kikuchi, V. Testa, M. Telles, J, Pereira, L. Dutra, and C. Sampaio 1999. First coral reef assessment in the southern hemisphere applying the AGRRA rapid protocol (Caramuanas Reef, Bahia, Brazil), International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring, and Restoration. National Coral Reef Institute, p. 112-123 (Abstract). Leao Z.M.A.N., R.K.P. Kikuchi, and V. Testa In press. Corals and coral reefs of Brazil. In: J. Cortes (ed.), Latin America Coral Reefs. Elsevier Publisher, New York. Martin L., J.M. Flexor, D.Blitzkow, and K. Suguio 1985. Geoid change indications along the Brazilian coast during the last 7,000 years. Proceedings of the Fifth International Coral Reef Symposium 3:85-90. Pitombo F., C.C. Ratto, and M.J.C. Belem 1988. Species diversity and zonation pattern of hermatypic corals at two fringing reefs of Abrolhos Archipelago, Brazil. Proceedings of the Sixth International Coral Reef Symposium 2:817-820. Zar, J. H. 1999. Biostatistical Analysis. Prentice Hall, New Jersey, 931 pp. Table 1. Site information for AGRRA stony coral and algal surveys in the Abrolhos, Brazil. 184 c o T3 CO m ro CO p O) o +1 — Tj* 40 40 (U C -H -H -H -H -H Ln C3 40 o 40 P O e rn vd 00 (N CO E i-> 'b o r4 Al O o c o O fO 5.5 10.5 CN CO <) <1> V5 O a u. a u« C43 JL t 3 s a, < C/5 O 'O O o 3 o a> (U Q!i a, X 0) a. X X X X X X ■4— • X) (U £ 03 £ o m (U T3 -o 03 'O C 03 +1 e 03 1) c o "O c o o •73 c 03 (U _N 25 cm diameter) by site in the Abrolhos, Brazil. 186 187 (U jc ■i— * .s 0) X5 s: <3 3 T3 c CO o D CO o o c o 4-1 o cn C 2 > (U T3 ^3 V-H CO "O c CO +1 C cO -D < 188 r-' ® 'SV2^-'SAB7- 38M2' W 0 /' sab5t: !! j v'ftVdbnda Is; 0:SAJ^Q ,_Guarita ' '\0 Is. ^t. Barbara Is. ^ Coral reefs Sand shoal o Lighthouse (8) AGRRA sites 0 0.4 km . • // # '' ' .5m •#*»<. suestels. Figure 1. AGRRA survey sites off Santa Barbara Island (SAB) and in the Parcel dos Abrolhos chapeiroes (PAB), Abrolhos, Brazil. See Table 1 for site codes. RAPID ASSESSMENT OF THE ABROLHOS REEFS, EASTERN BRAZIL (PART 2: FISH COMMUNITIES) BY RUY K. P. KIKUCHI,’’^ ZELINDA M. A. N. LEAO,^ CLAUDIO L. S. SAMPAIO,^ and MARCELO D. TELLES^ ABSTRACT The first rapid assessment of fish communities in the Abrolhos Marine National Park was undertaken in three offshore chapeiroes and at eight sites in the fringing reef off Santa Barbara Island. Fish communities in the two habitat types differed at both the family and species levels. Total species numbers were higher in the chapeiroes (38- 40/site) than in the fringing reef (13-37/site). Total fish density was higher in the fringing reef than in the chapeiroes, as were the densities of most surveyed families, except for pomacentrids, lutjanids and scarids. The most common size class for key herbivores (scarids >5 cm, acanthurids) and carnivores (lutjanids, large-sized serranids) was 11-20 cm. Herbivores grazing dead skeletal areas may accidentally damage stony corals, particularly in the chapeiroes. Limited impacts from sewage discharge may affect the northern coast of Santa Barbara Island; tourists or artisanal fishers may have reduced the density of carnivores in this fringing reef INTRODUCTION Brazilian coral reefs, which are the southernmost reefs in the western Atlantic o o Ocean, extend approximately 2,000 km between 0 50’ S and 19 S latitude. Although they have been studied for over a century, many reef areas are still poorly known and there are few quantitative assessments of their condition (see Leao et al., 1999). The Abrolhos reef complex in the southern part of the eastern state of Bahia (between 17°S and 19°S) contains the largest and most thoroughly investigated coral reefs in the western South Atlantic (Fig. 1). The reefs are scattered over an area of approximately 6,000 km^ on the inner and middle continental shelf of the Abrolhos Bank. There are three reef ' Former address: Universidade Estadual de Feira de Santana, BRl 16, Km 3, Campus Universitario, Feira de Santana, 44031-460, Bahia, Brazil. Universidade Federal da Bahia, Laboratorio de Estudos Costeiros, Rua Caetano Moura 123, Salvador, 40210-340, Bahia, Brazil. Email: rkikuchi@cpgg.ufba.br Universidade Federal da Paraiba, Departamento de Sistematica e Ecologia, Cidade Universitaria, Joao Pessoa, 58059-900, Paraiba, Brazil. 190 types: isolated nearshore bank reefs of varied shapes (e.g., elongated, circular) and sizes (<1 km to ~20 km) in depths of 10-20 m, offshore shallow (<10 m deep) fringing reefs that parallel the shorelines of the five volcanic islands of the Abrolhos Archipelago; and “chapeiroes”- huge (15 to >25 m high), isolated mushroom-shaped reefs found in water depths greater than 20 m. The northern coasts of the Abrolhos islands are exposed to relatively high wave energies between September and February, their southern sides to somewhat larger waves from April to August, while the western (leeward) coasts are protected from major wind trends. The fringing reef flats extend about 30 m from shore, are poorly developed, and, during low tides, are subaerially exposed. To seaward, the reefs gradually slope to their edges at depths of 4-5 m where they drop off to a sandy bottom at 8-10 m. The windward Parcel dos Abrolhos (PAB) chapeiroes are located about 2 km east of the Abrolhos islands (Fig. 1). Due to the limited penetration of light beneath the overhangs around their sides, the maximum development of stony corals is at depths of 5-8 m on the crests of the chapeiroes. A marine protected area was established in the Abrolhos reef complex in 1983 in recognition of its unique reef morphology, the presence of endemic scleractinians, and because most Brazilian reefs occur in this region. As the focus of the Abrolhos Marine National Park (AMNP) was to preserve the isolated chapeiroes, ninety percent is located approximately 60 km from the mainland in the offshore reefs, with only 10% of the Park in the coastal zone. Meanwhile, as urban expansion and deforestation have accelerated, the coastal reefs are being exposed to increasing amounts of river runoff and untreated sewage discharges. However, the offshore reefs are located beyond the reach of the turbid coastal waters. It is important to note that endemism of fish species in the Brazilian province is about 18% to 20% (Floeter and Gasparini, 2000, 2001) and the Abrolhos area is the southernmost geographic limit of some endemic tropical Brazilian species (e.g., Dasyatis marianne, Gomes et al., 2000; Moura et al., 2001). Almost 20 years have passed since the AMNP was established, yet there have been only a few assessments of its fish communities (Telles, 1998; Ferreira and Gon9alves, 1999). While there has been, to date, no evidence of excessive harvesting in the offshore reefs, which are designated as “no- take” zones, the coastal reefs appear to have been overfished (Ferreira and Gon9alves, 1999). Some of the fringing reefs that border the offshore Abrolhos islands have been intensively used by recreational divers and snorkelers since 1989 (Leao et al., 1994). Although tourists dive all around Santa Barbara Island (SAB) in particular, its southern side is most heavily visited. The PAB chapeiroes are less popular due to difficulty of navigation and rougher seas but its fish and coral populations are considered to be in relatively good condition (personal observations). Previous assessments of potential human disturbances in the AMNP have only referred to the anchorage of tourist boats in algal beds around the islands (Creed and Amado Filho, 1999). Our initial purpose in using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocols in the AMNP was to collect information on the condition of the offshore chapeiroes and the SAB fringing reef in order to establish standards for a long- term monitoring program in the park. The results of the benthos assessment can be found in Kikuchi et al. (this volume). 191 METHODS Fish surveys were conducted in eight fringing reef sites around SAB which are considered representative of its northern (three sites), southern (three sites) and western (two sites) coasts. The three southern sites were also strategically chosen because they are preferred recreational dive sites. Each site was located in shallow (3. 5-5. 5 m) water as close as possible to the seaward edge of the reef The 6-6.5 m crests of three representative chapeiroes were surveyed. Except for the leeward fringing reef, most of these sites are exposed to large waves, and the leeward sites experience strong tidal currents. On each dive, two divers utilized the AGRRA Version 2.0 fish protocol (see Appendix One, this volume), which was modified as follows. As many of the chapeiroes are less than 30 m in width, we made 30 transects/site, each 10 m long by 2 m wide. Except for the absence of a T-bar for estimating fish lengths, the transects were swum in the AGRRA manner. Juvenile grunts (haemulids) and parrotfishes (scarids) less than 5 cm in total length were tallied, as were all species of common reef fishes. Any juvenile that could not be identified to species was only assigned a genus name. The name Scams trispinosus Valenciennes, 1840 was substituted for Scams coelestinus which is no longer thought to occur in Brazil (Moura et al. 2001). Other identifications were based on the descriptions of Humann (1994), Smith (1997), Carvalho-Filho (1999) and Rocha and Rosa (2001). Roving Diver Technique surveys were not made. Therefore, the number of fish species at each site is taken from the transect results. To ensure consistency, all analyses are based on the data collected by Sampaio. Juvenile grunts and parrotfishes <5 cm in length were excluded from calculations of AGRRA fish densities. Scatterplots and regression plots were used to investigate possible relationships between fish characteristics [total densities and densities of key carnivores, key herbivores, surgeonfishes (acanthurids), and parrotfishes] and some ecological parameters [live stony coral cover, height or partial-colony mortality of “large” (>25cm diameter) stony corals, macroalgal index = macroalgal relative abundance x macroalgal height] that are given in Kikuchi et al. (this volume). Comparisons were made with both reef types (SAB fringing reef and the PAB chapeiroes) grouped together and separately. Regression coefficients of determination and significance probability based on ANOVAs are presented for the best relationships. Densities of surgeonfishes, parrotfishes >5 cm, and both herbivores combined showed a rather good fit to a normal distribution model; hence fish densities and numbers of species were used without transformation. Cluster dendrogram and MDS diagrams were plotted based on Bray-Curtis similarity coefficients between the mean total density of fishes grouped by family at each site after the mean densities were square-root transformed. The complete linkage option was chosen in hierarchical agglomerative clustering (Clarke and Warwick, 1994). 192 RESULTS The total fish community at many (7/11) sites had over 30 species of common reef fishes, and three of the remaining sites had more than 20 species (Table 1). Total species numbers were consistently higher at 6.5 m in the chapeiroes (39-40/site, n=3) than at 3. 5-5. 5 m in the fringing reef (13-37/site, n=8). In contrast, 6/8 of the SAB sites had 17-19 of the species on the AGRRA list, whereas corresponding numbers in the PAB were somewhat lower (14-1 6/chapeirao). Fish densities (AGRRA species only) were higher off SAB than in the chapeiroes (Fig. 2, Table 2) as were densities of the numerically dominant grunts (all sizes) and surgeonfishes. When all sites are considered together, the small omnivorous grunt, Haemulon aurolineatum, had the highest density, followed by another omnivore (and non-AGRRA-listed fish) the sergeant major, Abudefduf saxatilis, damselfishes {Stegastes spp.), two surgeonfishes (Acanthurus chirurgus and A. coeruleus) and, in sixth position, the carnivorous snapper (lutjanid), Ocyurus chrysurus (Table 3). Densities of parrotfishes (>5 cm) were about 20 individuals/ 100 m in both habitat types (Fig. 2). Acanthurids were more plentiful off SAB (~60 individuals/ 100 m^) than in the chapeiroes (35 individuals/ 100 m^). Snapper densities in the PAB were double those in the fringing reef (30 versus 15 individuals/ 100 m^, respectively). Groupers were fewer than about 5 individuals/ 1 00 m everywhere. t (/) o z s: LU Q 100 80 60 □ SAB ■ PAB .f*Ti^.rSa A^ A^ ^ r 0^ A & FISH FAMILY Figure 2. Mean fish density (no. individuals/100 ± sd) for all species (and all sizes of fish) in 1 1 fish families in the SAB fringing reef and PAB chapeiroes, Abrolhos, Brazil. Other = Caranx ruber, Caranx crisus, Sphyraena barracuda. 193 Stress=0.05 Figure 3. (A) Hierarchical cluster analysis and (B) MDS ordination plot of all common reef fishes (based on belt transect data), by site in the Abrolhos, Brazil. Site comparisons at the family level clearly separated the fringing reef from the chapeiroes (Fig. 3) on the basis of both the Bray-Curtis similarity matrix and the MDS analyses. The size-frequency distributions of the key carnivores (snappers and large groupers) were similar in the two habitats (Fig. 4A), the most common length class being 1 1-20 cm. The mean size of the key herbivores, parrotfishes >5 cm, and surgeonfishes (the damselfish Microspathodon chrysurus was not seen despite its known presence in Brazilian waters) was a little larger in the chapeiroes than in the fringing reef, even though the most common length class at both was 1 1-20 cm (Fig. 4B). Carnivore sizes in both environments (fringing reef and chapeiroes) were normally distributed, as were herbivores in the chapeiroes, however, the size distribution of the SAB herbivores was slightly skewed towards larger fishes. When all sites are included in the analysis, regardless of whether or not juvenile scarids (<5 cm in length) are retained, herbivore density showed no relationship to macroalgal index (Fig. 5 A). However, when the two island sites with the highest macroalgal indices (SABS, SABS) and the chapeiroes reefs are removed, herbivore (scarids >5 cm long, acanthurids) density was inversely related (R =0.800, p=0.016) to the macroalgal index of the remaining six SAB sites (Fig. SB). A separate test of surgeonfish density versus macroalgal index in these six sites also produced good fit (R =0.910, p=0.003), but there was no relationship between parrotfish (>S cm) density and macroalgal index in these same six sites. Despite the moderately good fit between herbivore density and macroalgal index in the chapeiroes, the significance test indicates a high probability that the curve is horizontal (R^=0.619, p=0.423). The total number of fish species/site showed no relationship with the density of large (>25 cm in diameter) stony corals when all sites are taken into consideration 194 (Fig. 6), and the positive regression for the chapeiroes sites is not significant (R^=0.881, p=0.224). No relationship emerged from the comparison of live stony coral cover and fish density. Neither was fish density related to the density of large stony corals nor to their average maximum height (which was used to approximate the rugosity of the reef surface, Kikuchi et al., this volume). Figure 4. Size frequency distribution of (A) carnivores (all lutjanids, select serranids) and (B) herbivores (all acanthurids, scarids >5 cm) in the SAB fringing reef and PAB chapeiroes, Abrolhos, Brazil. When all sites are considered, no relationship was found between the total density of the AGRRA fishes and values for old partial-colony mortality of the large stony corals. Amongst the herbivores, only in the chapeiroes (Fig. 7A, B, C) was old (but not recent) partial mortality positively related to the density of parrotfishes (R^=0.978, p=0.031) and surgeonfishes (R =0.997, p=0.037). DISCUSSION The structure of the SAB reef fish community seems representative of the offshore insular fringing reefs in the Abrolhos Archipelago as a whole. Our results are very similar at the family level to those of Telles (1998) who, in 1995, quantified 20 common fish taxa in sites that were evenly distributed among the fringing reefs of all five islands. (Four of these sites were located in SAB, three were in Sueste, two each were in Redonda and Siriba and one was in Guarita.) Telles’ (1998) belt transects, which were 50 m X 4 m, each being surveyed three times (i.e., there were three replicates of each 200 m^ transect), were swum at depths of 2-6 m. As in our SAB surveys, grunts, surgeonfishes and damselfishes were the most common families found by Telles (1998) when the data from all five islands are combined (Fig. 8). The SAB fish community also seems to be representative of the archipelago’s five fringing reefs at the species level, given the similarities in the relative frequency of Telles’ (1998) 20 target taxa in the two studies (Table 4). Haemulon aurolineatum is the most abundant of these species in both surveys. 195 Figure 5. Regression plots between mean herbivore density (no. individuals/100 m^) and macroalgal index (A) by site, (B) for six of the SAB sites (with the significant regression line), in the Abrolhos, Brazil. Figure 6. Regression plot between total fish species number and mean density of stony corals (> 25 cm diameter)/! 0 m, by site in the Abrolhos, Brazil. Regression line shown for the PAB chapeiroes sites is not significant. 196 C 100 T 80 - ^ 60 - o E 20 0 ♦ ♦ SBI ■ PAB — Linear (PAB) 0 20 40 60 80 100 120 140 Acanthurid density (#/100 m^) Figure 7. Regression plots between old partial colony mortality of stony corals and density of (A) key herbivores, (B) scarids >5 cm, (C) acanthurids, by site in the Abrolhos. Regression lines shown for the PAB chapeiroes sites are significant. 197 However, there are marked differences, at least at the family level, between the fish communities of the SAB fringing reef and the PAB chapeiroes (Figs. 2, 3). The positive (but non-significant) relationship between the number of fish species and the density of large stony corals in the chapeiroes might be explained by an attendant increase in the variety of habitats and niches associated with the latter. Like the large chapeiroes, massive colonies of Mussismilia braziliensis are mushroom shaped. Their upper (illuminated) surfaces are larger in diameter than their shaded bases, which are dead and overgrown by crustose coralline algae, turf algae, macroalgae, and sponges. As the number of colonies increases, concomitant increases in the microtopographic complexity may enhance fish species richness by increasing the number of ecological interactions among stony corals, other benthic species, and fishes. FISH FAMILY Figure 8. Percent relative frequency (number of fishes belonging to each family divided by the total number of fish counted) for 20 common fish taxa, by familiy, in the SAB fringing reef (this study) and all five offshore fringing reefs (Telles 1998) in the Abrolhos, Brazil. Other = Caranx ruber, Caranx crisus, Sphyraena barracuda. Local nutrient enrichment is possibly responsible for the unexpectedly high algal biomass at the two fringing reef sites in which the inverse relationship elsewhere found between herbivores and macroalgae at SAB is lacking (Fig. 5). One of these sites (SAB5) is located near an area known as Bird’s Cemetery where dead bodies of masked booby {Sula dactylatra), brown booby {Sula leucogaster), and magnificent frigatebird {Fregata magnificens) are frequently found. The other site (SABS) is seaward of the seven houses occupied by Navy and Park personnel where sewage either seeps from septic tanks or is discharged directly into the ocean. 198 Relative to the chapeiroes, snapper densities in the SAB fringing reef are strikingly low (Fig. 2) and the size-frequency distributions of herbivores are shifted toward smaller individuals (Fig. 4). As these patterns are not what we would expect in an area lacking direct human exploitation, tourists may simply chase carnivores away and/or illegal fishing (by tourists or artisanal fishermen) may occur at night. Recent partial mortality of large (>25 cm in diameter) stony corals was rare and bore no relationship to fish densities. In the chapeiroes, however, there seems to be a relationship between the densities of parrotfishes and surgeonfishes and old partial- colony mortality. Herbivorous fishes graze intensively on the dead surfaces of stony corals and may occasionally destroy some live tissues. Schools of the endemic greenbeak parrotfish, Scarus trispinosus, and of surgeonfishes are commonly seen engaged in this activity (personal observations of all co-authors). The species richness of the common and AGRRA-listed reef fishes and the sizes of key herbivores and carnivores was higher in the PAB than off SAB, reinforcing our interpretation from the benthos assessment that the chapeiroes are in better condition overall than the fringing reef (Kikuchi et ah, this volume). The role of tourism in depressing carnivore, and possibly parrotfish, density in the SAB fringing reef can only be tested by assessing other areas of the AMNP, such as the islands of Sueste and Guarita, where tourist activities are prohibited. If Telles’ (1998) data are any guide, carnivorous snappers and groupers are slightly more common in the five islands as a whole than off SAB. Anthropogenic effects are possibly reflected in the AGRRA fish indicators in what is supposed to be a protected fringing reef To test this hypothesis is one of the aims of future AGRRA assessments in the AMNP. We also plan to survey fish communities in its coastal reefs where artisanal and recreational harvesting is known to be intense. In sum, park management practices should be evaluated, and some changes may be needed to conserve the endemic species and other fishes in the Abrolhos Marine National Park. ACKNOWLEDGMENTS The authors acknowledge the logistical support offered by the Abrolhos National Marine Park during the fieldwork as well as the financial support by the Geology Graduate Course of Federal University of Bahia. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. We are thankful to Judith Lang and other Organizing Committee members for assistance with the revision of this article. REFERENCES Carvalho-Filho, A. 1999. Peixes da costa brasileira. Sao Paulo, Editora Melro, 318 pp. Clarke, K.R., and R.M. Warwick 1994. Change in marine communities: an approach to statistical analysis and interpretation. Plymouth, Natural Environmental Research Council, UK. 144 p. 199 Creed, J.C., and G.M. Amado Filho 1999. Disturbance and recovery of the macro flora of a seagrass {Halodule wrightii Ascherson) meadow in the Abrolhos Marine National Park, Brazil: an experimental evaluation of anchor damage. Journal of Experimental Marine Biology and Ecology 235:285-306. Ferreira, C.E.L., and J.E.A. Gon9alves 1999. The unique Abrolhos reef formation (Brazil): need for specific management strategies. Coral Reefs 18:352. Fleeter, S.R., and J.L. Gasparini 2000. The southwestern Atlantic reef-fish fauna: composition and zoogeographic patterns. Journal of Fish Biology. 56:1099-1 1 14. Fleeter, S.R., and J.L. Gasparini 2001. The Brazilian endemic reef fishes. Coral Reefs 19:292. Gomes, U.L., R.S. Rosa, and O.B.F. Gadig 2000. Dasyatis marianae sp.n.: A new species of stingray (Chondrichthyes: Dasyatidae) from the Southwestern Atlantic. Copeia (2): 5 10-5 15. Humann, P. 1994. Reef Fish Identification: Florida, Caribbean, Bahamas, 2"^. Ed. Vaughan Press, Orlando, Florida. 426 pp. Moura, R.L., J.L. Figueiredo, and I. Sazima 2001. A new parrotfish (Scaridae) from Brazil, and revalidation of Sparisoma amplum (Ranzani, 1842), Sparisoma frondosum (Agassiz, 1831), Sparisoma azillare (Steindachner, 1878) and Scarus trispinosus Valenciennes, 1840. Bulletin of Marine Science 68:505-524. Leao, Z.M.A.N., M.D. Telles, R. Sforza, H.A. Bulhoes, and R.K.P. Kikuchi 1994. Impacts of tourism development on the coral reefs of the Abrolhos area, Brazil. Pp. 254-260. In: R.N. Ginsburg (compiler) Global Aspects of Coral Reefs - Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, Miami. Leao, Z., R. Kikuchi, V. Testa, M. Telles, J. Pereira, L. Dutra, and C. Sampaio 1999. First coral reef assessment in the southern hemisphere applying the AGRRA rapid protocol (Caramuanas Reef, Bahia, Brazil). International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring, and Restoration. National Coral Reef Institute, p. 122 (Abstract). Rocha, L.A., and R.S. Rosa 2001. Halichoeres brasiliensis (Bloch, 1791), a valid wrasse species (Teleostei: Labridae) from Brazil, with notes on the Caribbean species Halichoeres radiatus (Linnaeus, 1758). Aqua, Journal of Ichthyology and Aquatic Biology 4(4): 161-166. Smith, C.L. 1997. Field guide to tropical marine fishes of the Caribbean, Gulf of Mexico, Florida, the Bahamas and Bermudas. New York, Alfred A. Knopf, 470 pp. Telles, M.D. 1998. Modelo trofodindmico dos recifes em franja do Barque Nacional Marinho dos Abrolhos - Bahia. Master Thesis, P6s-Gradua9ao em Oceanografia Biologica. Rio Grande, Funda9ao Universidade do Rio Grande, 150pp. Table 1. Site information for AGRRA fish surveys in the Abrolhos, Brazil. 200 Table 2. Density (mean ± standard deviation) of AGRRA fishes and macroalgal index, by site in the Abrolhos, Brazil 201 202 Table 3. Twenty-five most frequently sighted fishes and the mean densities for species seen in belt transect surveys in the Abrolhos, Brazil. Scientific name Sighting frequency (Vof Density (#/100 m^) Abudefduf saxatilis 100 28.95 Stegastes spp. 100 23.77 Acanthurus chirurgus 100 20.06 Acanthurus coeruleus 100 17.79 Ocyurus chrysurus 100 16.88 Scarus trispinosus' 100 11.77 Sparisoma spp. 100 7.9 Anisotremus virginicus 100 4.89 Chaetodon striatus 100 3.33 Holocentrus adscensionis 100 2.21 Acanthurus bahianus 96 13.77 Pomacanthus paru 96 2.98 Mycteroperca bonaci 96 1.86 Gramma braziliensis 96 1.59 Haemulon plumieri 92 5.32 Balistes vetula 88 0.83 Haemulon aurolineatum 83 35.68 Halichoeres radiatus 79 1.61 Elacatinus figaro 75 5.68 Holocanthus ciliaris 75 1.05 Haemulon parra 71 9.73 Scarus zelindae 71 1.48 Caranx chrysus 67 1.36 Caranx ruber 50 0.86 Caranx bartholomaei 46 1.02 formerly Scarus coelestinus. ^Sighting frequency (%) = percentage of dives in which the taxon was recorded. 203 Table 4. Relative frequency of the most common fishes in the Abrolhos Archipelago (Telles, 1998) and in Santa Barbara Island (this survey). Scientific name Relative frequency (%) (Telles, 1998) (this survey) Haemulon aurolimatum 19 14 Stegastes variabilis 13 9 Acanthurus chirurgus 11 8 Scarus trispinosus' 8 4 Abudefduf saxatilis 7 11 Chaetodon striatus 5 1 Caranx latus 5 0 Ocyurus chrysurus 4 7 Acanthurus coeruleus 4 7 Pomacanthus arcuatus 3 0.2 Anisotremus virginucus 3 2 Haemulon parra 3 4 Sparisoma spp. 3 3 Mycteroperca bonaci 3 1 Batistes vetula 3 0.3 Lutjanus jocu 2 0.3 Caranx rubber 2 0.3 Holocentrus adscensionis 1 1 Kyphosus sectatrix 1 4 Sphyraena barracuda 1 0.1 formerly Scarus coelestinus. ^Relative frequency = number of individuals/taxon relative to the total number of individuals. 204 Figure 1. AGRRA survey sites in Grand Cayman and Little Cayman, Cayman Islands. See Table 1 for site codes. STATUS OF CORAL REEFS OF LITTLE CAYMAN, GRAND CAYMAN AND CAYMAN BRAC, BRITISH WEST INDIES, IN 1999 AND 2000 (PART 1: STONY CORALS AND ALGAE) BY CARRIE MANFRINO,* BERNHARD RIEGL,^ JEROME L. HALL,^ and ROBERT GRAIFMAN^ ABSTRACT A benthic assessment of the isolated Cayman Islands was completed at 42 sites. Major changes in the reef community structure were documented by comparison with earlier studies. Acropora palmata and A. cervicornis, once abundant as shallow framework builders, were uncommon. Diseased stony corals were seen in >90% of the study sites, with the highest averages in Little Cayman, especially at Bloody Bay which is one of the most highly regulated marine parks in the Cayman Islands. The Montastraea annularis species complex accounted for two-thirds of the diseased corals which, along with other massive species, were affected largely by white-plague disease. Recent partial-colony mortality was particularly high in Grand Cayman. However, small- to intermediate-sized (<1.5 m diameter) colonies and recruits of reef framework builders (including the M annularis complex) suggest a strong potential for population regeneration. Algal competition generally did not appear to be a problem for stony corals, and bleaching was insignificant, yet more prevalent, in the deeper (>10 m) sites. INTRODUCTION The Cayman Islands lie in the middle of the Caribbean Sea, about 240 km south of Cuba. They are comprised of three small (12-18 km long) low-lying, limestone islands with Little Cayman (EC) 125 km northeast of Grand Cayman (GC) and just 7.5 km to the southwest of Cayman Brae (CB). EC has a permanent resident population of about 130, and remains relatively undeveloped with four small dive *Kean University, Department of Geology and Meteorology, 1000 Morris Ave. Union, NJ, 07083. Email: cmanfrin@kean.edu National Coral Reef Institute, Nova Southeastern University, 8000 N. Ocean Drive, Dania Beach, Florida, 33004. ^Institute of Nautical Archeology, Texas A & M University, College Station, TX, 77843. '^Marine Environmental Education and Research Institute, P.O. Box 1461, Princeton, NJ. 206 lodges and one 40-room resort. There are 2,200 residents and three dive resorts on CB. In comparison, GC has a resident population of 36,000, and numerous large resort hotels, condominiums, and dive operators along the seven-mile beach area on the western side of the island. Georgetown Harbor is an important port for cruise ships. The Cayman Islands are protected from open-ocean long-fetch Atlantic waves by the Antilles Island Arc, Hispaniola, and Cuba. Trade winds blow predominantly from the northeast but shift periodically to the east and southeast. The eastern side of GC receives the highest wind and wave energy, whereas the western sides of all three islands are least exposed. LC and CB are oriented northeast to southwest, so the southern sides tend to be more exposed to prevailing winds and waves. Polar fronts push winter storms south during the winter months, and winds around the islands shift to blow from the north, developing into Noreasters and Norwesters. Two shallow-sloping submarine terraces occur around the islands with a few exceptions (e.g. Bloody Bay on LC is missing the lower terrace) (Rigby and Roberts, 1976; Blanchon et al., 1997). The upper terrace extends from the shoreline to a mid- shelf scarp at depths of 8-15 m and a lower terrace from the base of this scarp to the edge of the shelf at 15-20 m. Spectacular shelf-edge “walls” occur around all three islands. Fringing reefs and boulder ramparts are nearly continuous around LC and GC (except off the western side of GC) and occur along the south side (western third) of CB. In many locations, the boulder rampart (in 0.5-5 m water depth) was formerly a reef crest dominated by Acropora palmata on the upper terrace along the windward coasts (east, south, north of GC and LC, south of CB). High-relief (to 12 m) spur-and- groove or buttress reefs have developed at several sites at the edge of the upper terrace in the more exposed (eastern, northeastern, southeastern) sides of the islands. At “more protected” windward locations on the northern and southern coasts, low- to medium-relief (1-3 m) spurs and grooves are found at the edge of the upper terrace. In leeward locations (western side of GC; Bloody Bay on LC), where a reef crest is lacking, there are narrow, low-to-medium relief spurs and grooves or shelf-edge reefs with poorly developed spurs and grooves, although a few high-relief, elongated spurs occur along the upper terrace in Bloody Bay. Episodes of coral bleaching were reported in the Cayman Islands in the late 1980’s, 1995, and in 1998 when nearly 80% of the corals bleached on GC (LC and CB were not surveyed) (Timothy Austin, personal communication). The last major storms to impact the Caymans were Hurricanes Allen in 1980, Gilbert in 1988, and Mitch in 1998. Hurricanes Allen and Mitch reportedly had minimal impact underwater but Gilbert damaged the shallow reefs, particularly breaking small colonies of Acropora palmata, at many locations around GC (Blanchon and Jones, 1997; Timothy Austin, personal communication). The extent of the impacts by these storms remains largely undocumented on LC and CB; however, boulder ramparts on many of the beaches of all three islands and reef-crest zones that are largely biodetrital provide historic evidence of the continuous impact by major storms. A Marine Conservation Law was passed in the Cayman Islands in 1979. Marine park areas (15 km^), replenishment zone areas (15 km^), an environmental zone (17 km^) in the North Sound of GC, and animal sanctuaries/RAMSAR sites were designated in 1986 (Fig. lA, B). Conservation regulations have not prevented 207 overfishing of conch and lobster, nor have they adequately protected grouper spawning grounds from fishing or turtle nesting sites from coastal development. Residents (for at least five years) can apply for licenses to use spearguns and seine nets and to capture turtles if they can prove these activities are a cultural necessity. A total of 500 speargun licenses have been issued in the Cayman Islands. Growth in the hotel industry has resulted in an influx of workers from other Caribbean countries and has resulted in increased use of fish traps (pots). Recent amendments made to the Marine Conservation Legislation in the Cayman Islands include instituting a season for conch and limiting the catch to five per person (from 15) and 10 per boat (from 20). Grouper spawning sites are now protected, and a catch limit of 12 per boat and a minimum size of 12 inches have been set. Individuals normally resident in Cayman are permitted to line fish in these areas. Lobster catch limits have been reduced to three per person and six per boat per day, and the length of the closed season has been extended. Fish pots are now limited to a maximum of two per household, and they must now be licensed and tagged. Only Caymanians are granted fish pot licenses. Beginning in the early 1980’s, the Cayman Islands Department of Environment has installed 159 permanent moorings on GC, 56 on LC, and 54 on CB. Diving has historically been focused along the western and northwestern coasts of GC where some dive sites receive over 15,000 visitors per year. Most divers on LC visit only one location. Bloody Bay Marine Park, where there are 22 moorings over a 3 km distance. With increased human usage (fishing, coastal development, mangrove destruction, diving, etc.), concern about the survival of Cayman’s coral reefs has been growing. The survey reported here was designed to rapidly assess the present status of these important resources with the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocols (Manfrino, et al., 2000) and to determine recent changes in reef community structure and condition in the various oceanographic settings by comparison with earlier accounts (as summarized by Rigby and Roberts, 1976; Roberts, 1994; Logan, 1994; Hunter, 1994). Evaluating the current distribution of Acropora palmata and A. cervicornis was a major concern because of the extensive distribution previously reported. Other objectives were to compare reefs off the more populous GC with those of its sister islands and to strategically survey sites both inside and outside of marine protected areas and those with both high diver pressure and low or no diver pressure. The Cayman Islands lie in the middle of the Caribbean Sea, far from most continental and many human influences, and thus should provide an important point of comparison for other AGRRA surveys throughout the region. A detailed report on the results of the fish assessments completed during this study can be found in Pattengill-Semmens and Semmens (this volume). METHODS A total of 1 1 scientists and graduate assistants conducted AGRRA benthos surveys on LC and GC in June 1999 and on CB in June 2000. As relatively little information existed for LC, its major reef types were initially characterized during a reconnaissance around the entire island using a standard manta-towing technique. We 208 recorded locations where coral cover was high and where we could potentially survey typical reefs in the recommended water depths for the AGRRA protocol. Except for the southwest coast of GC, the representative sites that were studied on all sides of the islands were strategically located both inside and outside of marine parks and replenishment zones. At LC there was a higher concentration of sites in heavily dived areas (Bloody Bay, Jackson Point). Published maps and aerial photographic images made available from the Cayman Islands Department of Environment 1994 photo series and input from knowledgeable boat captains were essential aides in site selection. All scientists participated in two days of training to standardize measuring techniques and identification of species. Humann (1993) was our primary identification guide. Following AGRRA protocol Version 2.2 guidelines (see Appendix One, this volume), two zones were examined: the shallow (1-5 m) Acropora palmata reef crest, and the generally well-developed framework reefs in depths of ~5-16 m on the upper terrace. Acropora palmata in the reef-crest zone was sparse, or completely dead, and although several sites were investigated, only one (CIl 1) was surveyed. Forty-one surveys were conducted on the upper terrace. The minimum size of surveyed stony corals was 10 cm and coral size was recorded to the nearest 10 cm. Agaricia was not identified to the species level. Damselfish in the vicinity of the surveyed corals were counted. All tiny corals (<2 cm) were included in the counts of “recruits.” Sediment was removed from crustose coralline algae before estimating their abundance in the quadrats. Algal abundances at CB were estimated using the AGRRA Version 3.1 protocols given in Appendix One, except they included turf algae. For the purposes of analysis, the upper terrace sites were arbitrarily divided by depth into 18 “shallower” reefs that were located in ^10 m and 23 “deeper” reefs that were >10 m deep. Few geomorphological or habitat differences were associated with these depth divisions. RESULTS A total of 5,53 1 stony corals (all >10 cm in diameter) in 486 transects and 2,407 algal quadrats were assessed during surveys off LC (18 sites) and GC (15 sites) in June, 1999, and off CB (9 sites) in June, 2000 (Table 1). Average live stony coral cover varied between ~14% in CB and 23% in LC (Table 1). Stony Coral Species Thirty-three scleractinian and hydrocoral species with colonies of at least 10 cm diameter were encountered along the transect lines. However, roving surveys indicated a total of 37 species in the vicinity of the survey sites (including species characteristically not growing larger then 10 cm). Reef crests. Acropora palmata, the once-dominant shallow reef-crest coral (Rigby and Roberts, 1976), was found only in rare patches just seaward of what is now a rubble rampart. Extensive groves of “standing dead” colonies (completely dead 209 but still in growth position) and “stumps” occurred in the windward reefs and off the north sides of GC and LC. Only one windward site on LC (CIl 1) was found with a zone of live A. palmata that was extensive enough to survey. Upper terrace reefs. Stony corals (>10 cm in diameter) off all three Cayman Islands were dominated by the Montastraea annularis species complex and Agaricia spp. (Fig. 2 A). Inter-island comparisons indicate that Montastraea annularis, M. faveolata and M franks i were numerically predominant in LC (>38%) and GC (~28%), bui Agaricia was predominant in CB (Fig. 2B). The Montastraea annularis complex generally dominated the shallower reefs while Agaricia spp. were predominant in the deeper reefs (Fig. 3A, B). AG 18% B Inter-island comparisons MA MAP MFR AG MC PP PA SS DIP MIL AP AC CN Other Coral Species Figure 2. (A) Species composition and mean relative abundance of all stony corals (>10 cm diameter) in all survey sites off the Cayman Islands. (B) Interisland comparison of 14 stony coral species. AC = Acropora cervicornis, AP = A. palmata, AG = Agaricia agaricites, CN = Colpophyllia natans, DIP = Diploria spp., DL = D. labyrinthiformis, DS = D. strigosa, MIL = Millepora spp., MA = M annularis, MC=M cavernosa, MAF= M faveolata, MFR = M franks i, PA = Porites astreoides, PP = P. porites, SS = Siderastrea siderea. 210 Figure 3. Species composition and mean relative abundance of all stony corals (>10 cm diameter) in (A) shallower (SIO m), (B) deeper (>10 m) upper terrace reefs off the Cayman Islands. See Figure 2 for species codes. A. palmata was present in five of the six most exposed windward reefs off GC (Cl 19, CI20, CI21, CI29, Cl 31) and in one protected windward reef along its northern coast (Cl 24). It occurred in three of seven windward reefs off LC (CIS, Cl 1 1, Cl 15) and four of five windward reefs off CB (CB2, CB4, CBS, CB6). Live corals were generally less common than standing dead colonies. Although thickets of Acropora cervicornis were seen more commonly landward of the reef crests, at least a few small (<1 m diameter) colonies were present in the shallower parts of the spur- and-groove formations at most (10/15) reefs off GC. A. cervicornis was rare in LC and CB. Piles of broken branches and dead, intact thickets were found in spur-and- groove zones, on hardgrounds, and generally at locations where some live colonies occurred. Stony Coral Size The average diameter of the >10 cm stony corals (Table 2) was essentially identical off LC (~45 cm) and GC (~42 cm) but smaller off CB (~32 cm). Average coral height was somewhat greater off LC (nearly 30 cm), where Montastraea faveolata was numerically more abundant than off the other two islands (Table 2). The largest corals were Montastraea faveolata (with heights as great as 320 cm and diameters up to 400 cm), but there were also large (up to 390 cm tall) colonies of Colpophyllia natans, Diploria clivosa, M. annularis, and Acropora palmata. The diameters of live A. palmata, however, were all smaller (<1.5 m) than the standing dead colonies (1.5-2 m in diameter). Surveyed M annularis occurred most frequently in the 21-60 cm diameter size range (Fig. 4 A) and colonies that were more than 1 m in diameter were more common in LC and GC than in CB. Colonies of Agaricia spp. 211 were most abundant in the 21-30 cm class size, and rarely exceeded 70 cm in diameter (Fig. 4B). No correlation between coral size (height or diameter, all >10 cm in diameter) and water depth or level of exposure to prevailing winds was identified (Tables 2 and 3). ^ Montastraea annularis complex Size class (diameter, cm) B tfi 'c 3. "o CJ tt-. o « S 3 :z Agaricia spp. <20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 Figure 4. Size-frequency distribution of colonies (>10 cm diameter) of (A) Montastraea annularis, M. faveolata and M. franksi, (B) Agaricia spp. in shallower (SIO m) and deeper (>10 m) upper terrace reefs off the Cayman Islands. Stony Coral Diseases and Bleaching Diseased stony corals were seen in over 90% (38/42) of the surveyed reefs (Table 2). Overall, the lowest incidence was at CB (mean = 2.3% diseased in 2000), while LC had the highest (mean = 5.5% diseased in 1999). Incidences of disease were 212 higher off the leeward sides of all three islands, affecting an average 6,5% of the leeward colonies versus 1.5 % of those off the windward coasts (Table 2). Heavily dived areas (Cl 8,9,10, 22, 23, 27, 32, 33, CB 6,7) had high disease frequencies. For example, the mean percentages of diseased corals were more than double at two of the most popular dive sites (Cemetery Reef and Sunset House) on the western (leeward) side of GC than on its other coasts (Table 2). The highest percentage (mean >24%) of diseased corals was found at the Meadows site in Bloody Bay Marine Park off LC, and the average percentage of diseased colonies in Bloody Bay (10.5%) was nearly double that of the island as a whole. More than 85% of these diseases were white syndromes (sensu Peters, 1997), i.e., white plague (WP) (presumably Type I, as subsequent visits indicated the disease was moving slowly across the coral tissues), which affected most species of massive A % Recent mortality 0-10 11-20 21-30 31-40 41-50 51-60 61-70 71-80 81-90 91-100 % Old mortality Figure 5. Frequency of (A) recent and (B) old partial colony mortality of stony corals (>10 cm diameter), by depth (^lOm and >10 m), in upper terrace reefs off the Cayman Islands. 213 corals except Montastraea franksi, and white-patch (white-pox) and white-band (WBD) diseases in Acropora (for descriptions of coral diseases see Peters, 1997). Black-band disease (BBD) and yellow-band (yellow-blotch) disease (YBD) only aceounted, on average, for 1 1.4% and 3.4%, respeetively, of the diseased corals off LC and GC, and neither were present in any of the transects or in random surveys off CB. No other diseases were noted in the study areas. Sixty-eight percent of the eorals affeeted by disease were in the Montastraea annularis speeies complex, whereas only 3% were species of Agaricia. YBD was only found in Montastraea annularis and M. franksi\ however, three colonies of Porites astreoides showed signs of a disease resembling YBD. BBD oecurred preferentially in species of massive corals. Regionally, the incidence of coral bleaching was low. The percent of stony eorals that were bleached averaged higher off LC (2.4%) and GC (1.7%) in 1999 than off CB (0.3%) the following year. Bleached corals were twice as common in the deeper upper-terrace reefs off all three islands as in the shallower reefs. The non- algal-gardening yellowtail damselfish (Microspathodon chrysurus) was the most abundant damselfish in the immediate vicinity of the surveyed stony corals. In LC, 141 individuals/2, 542corals (5.5%) were counted; corresponding values for GC were 84 fish/2,162 eorals (3.9 %), and 19 fish/954 corals (1.99%) for CB. Stony Coral Mortality Recent partial-colony mortality (hereafter reeent mortality) was in the 0-10% range for most stony corals (>10cm diameter). Recent mortality in GC in 1999 (5.0%) was more than double that in LC (2.1%); eorresponding values were even lower (1.3%) in CB in 2000 (Fig. 5A; Table 2). Overall, areas with the highest abundance of disease were also areas with high recent mortality. Old partial mortality (hereafter old mortality) averaged 21.3% of these eolony surfaces, and there were no significant depth-related trends or differences between the islands and no differences related to exposure (Fig. 5B; Table 2). Apart from the Bloody Bay Marine Park area, “standing dead” eorals were uneommon off LC, but they oceurred at all but two of the 15 GC reefs and all but one of the nine CB reefs (Table 2). Stony corals in about half of the GC (8/15) and LC (8/18) reefs but only a few (2/9) of those in CB had a total (= recent + old) partial mortality exeeeding 25% of their colony surfaces. Corals in leeward settings generally had higher overall mortality pereentages than those in more exposed windward orientations (Table 2). However, the only GC locations that had total mortality percentages of less than 20% were off the eastern end of the island. The highest percentages of reeent (10.5%) and total (36%) mortality were recorded at Isabel’s Reef, a relatively new dive site off the northern GC coast. High (>20%) to very high (>50%) percentages of reeently dead eolony surfaces in all three islands were most commonly found in Montastraea annularis, Diploria strigosa and D. labyrinthiformis. In the CB and LC reefs, Agaricia spp. had high pereentages of recent mortality. Additional species with high-to-very high pereentages of recent mortality off LC and GC included M. faveolata and M. cavernosa. In GC, Acropora cervicornis, Colpophyllia natans and Siderastrea siderea also commonly had high-to-very high percentages (10-60%) of recently dead 214 colony surfaces. On LC, recent mortality was also high (>20%) in M franksi (which was less abundant on CB and GC) and in Porites pontes. Stony Coral Recruitment Recruitment averaged 4.2 coralsW (= .26 ± .65 reeruits/.0625 m^), was patehy in all reefs, and followed no trends related to exposure to higher or lower wave energy (Table 4). Recruitment was slightly higher in GC than in LC and CB. The most eommon recruits in the upper-terrace reefs were Agaricia spp. and Porites astreoides (Fig. 6A, B). Although there was little difference overall in the species composition of the recruits at these two depths, Siderastrea siderea was nearly twice as common in the deeper sites as in the shallow sites. Small colonies of Montastraea annularis were found at all depths; recruits of M. faveolata were also present in the deeper reefs. Algal Relative Abundance The relative abundanee of turf algae was low, averaging 13.1% off CB to 22.4% off LC. The relative abundance of crustose coralline algae was highest in GC (mean = 53.5%) and lowest in CB (mean = 28%), whereas macroalgae showed the opposite pattern (30.5% versus 59%, respectively). Crustose coralline algal Figure 6. Species composition and mean relative abundance of all stony coral recruits (^2 cm diameter) in (A) shallower (^10 m) and (B) deeper (>10 m) upper terrace reefs off the Cayman Islands. AG = Agaricia agaricites, AC = Acropora cervicornis, AP = A. palmata, DIC = Dichocoenia, DC = Diploria clivosa, DL = D. labyrinthiformis, DS = D. strigosa, EF = Eusmilia fastigiata, FF = Favia fragum, MAN = Manicina areolata, MIL = Millepora spp., MA = Montastraea annularis, MC = M cavernosa, MFA = M. faveolata, PA = Porites astreoides, PP = P. porites, SS = Siderastrea siderea. 215 abundance was slightly higher in the deeper sites off GC and LC, while macroalgal abundance was higher on average in the deeper sites off all three islands. Turf algae were less abundant at the deeper sites in GC and LC. Because of taller growth, the average macroalgal index (relative abundance of macroalgae x macroalgal height) was nearly three times larger in CB than in GC (Table 4), and was also regionally higher in the deeper reefs. The highest macroalgal index value was found in a fairly remote reef (CB3) in a leeward/protected windward location on the northeast side of CB. Macroalgae were less abundant and macroalgal index values were lower in the more exposed windward sides of all three islands, except in LC where the macroalgal indices were slightly lower off the leeward side. The windward side of GC, however, clearly had the least macroalgae and the lowest macroalgal indices (which brought down the overall average for GC as well) (Tables 3, 4). Dictyota spp., Stypopodium zonale, Lobophora variegata, and Sargassum hystrix were the most common macroalgae, especially on the southern sides of the islands. In some reefs, macroalgal encroachment and overgrowth, with resultant mortality of stony corals, was observed. Diadema antillarum were only seen in transects at two reefs (one each shallower and deeper) on CB in 2000 and, even during roving dives, none were observed in the study areas off either GC or LC. DISCUSSION Comparisons with earlier studies of the coral communities in the Cayman Islands indicate that many changes have occurred over the past 30 years. The shallow Acropom palmata zone, as described for the 1970’s and 1980’s (Rigby and Roberts, 1976; Logan, 1994), no longer exists. In 1988, Hurricane Gilbert heavily impacted the north side of GC and destroyed much of the A. palmata-dommdXQd fringing reef (Blanchon et ah, 1997; T. Austin, personal communication). The predominance of standing dead colonies of A. palmata in many of the upper-terrace study sites on GC (Cl 19, 20, 21, 24, 29, 31) is indicative of prior, major mass mortality, possibly from white-band disease or, in part, from one or more of the earlier mass bleaching events. Tossed, dead fragments that were being consolidated by Millepora and crustose coralline algae, and a few small (<1 m in diameter) intact colonies of A. palmata, were all that remained in the shallow fringing reef crests on the north side of GC, where we had found higher densities during previous surveys in 1 997 (Manfrino, unpublished data). The colonies in this area, growing close to sea level, appear to be periodically smashed by the passage of major storms and were probably affected by Hurricane Mitch in 1998. Off all three islands, and at all depths and habitats in which they occurred, colonies of A. palmata and of A. cervicornis were affected by WBD and white-patch disease. Earlier studies indicate that WBD severely impacted A. cervicornis on the GC fore reef in the 1980s (Woodley et al., 1997). Nevertheless, these coral species appear to us to have the potential for strong regeneration in the Cayman Islands. Overall, since reefs with the highest abundance of disease were generally the reefs with high recent mortality, we presume that the latter was the result of diseases, particularly WP, WBB, and white patch. It is difficult to account for the observation 216 that LC had a higher percentage of diseased corals, but lower recent mortality values, than GC. One possible explanation is that outbreaks of disease in GC had occurred earlier that year, or in the year prior to our surveys, and was only at the time affecting the LC reefs. The higher occurrence of diseases in LC during June 1999 was possibly due to the higher percentage of Montastraea faveolata in the outer terrace reefs as WP had preferentially impacted the Montastraea annularis species complex. Surveys that continued on LC in 2000 and 2001 indicated outbreaks of disease activity with a trend of decreasing incidence of WP in 2001, then a major increase since January 2002 (further increasing in the 2002 summer months) that is causing a major mortality on the reefs around Little Cayman (Manfrino, unpublished data). The low occurrence of WP and low recent mortality (and the low live stony coral cover) on CB in the 2000 survey also supports a theory of waves of disease having previously swept through the CB reefs. Differences between the islands, between the different depths, and between the different island settings were documented. The islands differed in several ways. In CB, stony corals were smaller, live stony coral cover and recent mortality were lowest, as were incidences of disease and bleaching, but there was a higher percentage of standing dead coral than had been found the previous year off the other two islands. In GC there was higher recent and total mortality, higher abundance of crustose coralline algae, slightly higher coral recruitment, and lower macroalgal abundance and macroalgal index values (especially low off the windward GC coast). In LC, stony coral size and cover were greatest and though standing dead coral was lowest, the incidences of bleaching and disease were highest (especially in the Bloody Bay Marine Park where recent mortality was also high). Fish abundance was also higher in LC than in GC (see Pattengill-Semmens and Semmens, this volume). Disease and recent and total mortality were high in popular dive sites off GC and LC, which also happen to be in leeward settings. In general, mean coral diameter and height were greater in the shallower depths off all three islands. Diseased stony corals were more common but bleaching was less abundant in the shallower depths. The relative abundance of macroalgae was higher in the deeper reefs. No relationship between recruitment and level of exposure was found around any of the islands. Higher recruitment and coral regeneration off GC may occur because the higher abundance of crustose coralline algae in the windward reefs provides a more suitable surface for coral recruits. The difference in the overall condition of reefs within protected and unprotected areas was unexpected; stony corals in dive sites at Bloody Bay Marine Park, the most highly regulated park in the Cayman Islands, were undergoing a major mortality event in 1999, and this area had amongst the highest disease counts of all three islands. Although this survey did not include sites in any of the Cayman’s shelf-edge reefs, a limited number of wall dives made at Bloody Bay during June, 1999 indicated that bleaching was no longer significant in the deeper reefs. However, high levels of macroalgal overgrowth (up to 1 00%) on some of the stony corals suggested that a wave of recent mortality may have occurred in the aftermath of the 1998 bleaching event. 217 ACKNOWLEDGMENTS The project was coordinated by the Central Caribbean Marine Institute with the assistance and financial support of the Gilo Family Foundation, Kean University, Southern Cross Club, Little Cayman Beach Resort, Tortuga Divers, American Airlines and Cayman Airways. The AGRRA Organizing Committee facilitated financial support, as described in the Forward to this volume. The Austrian Science Foundation (FWF) grant PI 3 165-GEO, Reef Environmental Education Foundation (REEF), and Cayman Islands Department of Environment are recognized for their support. Our boat captains, Peter Hillenbrand, Buck Buckenroth, Jon Clamp and Vince Loengrant, generously shared with us their comprehensive knowledge about the reefs of the Cayman Islands. The scientific party to whom we are indebted for all their dedication and hard work included Christy Pattengill-Semmens, Brice Semmens, Casey Hermoyian, Leslie Whaylen, Belal Hansrod, and Marilyn Brandt. REFERENCES Blanchon P., and B. Jones 1997. Hurricane control on shelf-edge reef architecture around Grand Cayman. Sedimentology 44:9-506. Humann, P. 1993. Reef Coral Identification. Jacksonville, FL, New World Publications, Inc. 239 pp. Logan, A. 1994. Reefs and lagoons of Cayman Brae and Little Cayman Pp. 105-124. In: M. A. Brunt, and J. E. Davies (eds.). The Cayman Islands - Natural History and Biogeography. Kluwer Academic Publisher, Dordrecht, Netherlands. Manfrino, C., B. Riegl, C.V. Pattengill-Semmens, J.L. Hall, B. Semmens, K. Hoshino, R. Graifman, and C. Hermoyian 2000. Status of the reefs in the Cayman Islands. Ninth International Coral Reef Symposium, Bali, 2000, Indonesia, Abstracts, p. 359. Peters, E.C. 1997. Diseases of coral reef organisms. Pp: 1 14-129. In: C. Birkeland (ed.). Life and Death of Coral Reefs. C. Chapman and Hall, New York. Rigby J.K., and H.H. Roberts 1976. Grand Cayman. Island: Geology, Sediments, and Marine Communities. Brigham Young University Geology Studies 4:1-96. Roberts, H.H. 1988. Grand Cayman Swamps and Shallow Marine Substrates 1 and 2, Cayman Islands 1:25,000 (map). Overseas Development Natural Resources Institute, U.K. 218 Stoddart D.R., and M.E.C. Giglioli (editors) 1 980. Geography and Ecology of Little Cayman. Atoll Research Bulletin 241:1-180. Woodley, J.D., P. Alcolado, T. Austin, J. Barnes, R. Claro-Madruga, G. Ebanks-Petrie, R. Estrada, F. Geraldes, A. Glasspool, F. Homer, B. Luckhurst, E. Phillips, D. Shim, R. Smith, K. Sullivan-Sealy, M. Vega, J. Ward, and J. Wiener 2000. Status of coral reefs in the northern Caribbean and western Atlantic. Pp. 26 1-285. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Cape Ferguson, Queensland and Dampier, Western Australia. Australian Institute of Marine Science. Woodley, J.D., K. De Meyer, P. Bush, G. Ebanks-Petrie, J. Garzon-Ferreira, E. Klein, L. P.J.J. Pors and C.M. Wilson 1997. Status of coral reefs in the south central Caribbean. Proceedings of the Eighth International Coral Reef Symposium, Panama City. 1:357-362. Table 1 . Site information for AGRRA stony coral and algal surveys in the Cayman Islands. 219 >> a p p <0 o tn o to tn to to to o o > -H § ON NO (N o 00 tri wS NO NO 00 NO 00 O ■H -H -H ■H -H -H -H -H -H ± ± ± -H > O O o O w-j to O »/7 p o to to to to to o o E, (N eN r^‘ (N On On’ (N to tri , o o to . to , to to €0 >> C K 00 On Al c o w (A o o (A *5 c NO o m CN O NO o — > 1— > — i 0) o m O O _ 00 r- , 00 Tf ON Tl* m wn 7f 7f ON ro tT to 7f to "S) c O NO ii c o oc ii 73 U ii 73 ii 73 Di ii 73 c< ii 73 o c c a> C (1) Z4 C ii c ii (A c ii *5 C ii 17 c/5 CL Cu a. cx cx 3 CX ig 3 CL 3 3 3 CL CL o o CL CL CL o CL o o o 73 c5 73 o Cl o o o o o 73 73 -o 73 73 73 73 *o *0 73 "3 73 73 73 73 73 73 C c5 c5 c5 c5 e5 cd > 3 cd e3 c3 3 c3 c3 ig <3 :5 # ? ? $ # :S O a. ? :? -J 4> ou ■o OiJ 73 OJj 73 Oh 73 c c Oh O x: u Ln — ■ CN •— * »o ON 1—1 — — — < m NO — i/j U u u u u u u u u u u u u u u u u u U s f. © o JZ ii 3 2 UL c/5 a 5r H L- o 15 s: CX t/5 C3 w (A c ii 73 (A ii c a -o S ^ 2 u g a Q. 3 s o Paul's Ancho ii a C c c5 (n 73 3 H Lucas's Ledgi W 5 U 2 Site § i >3 Meadows S c/) £? 73 c 2 O Disneylan CQ (A 2 c5 JZ U No Name CX H o JS CQ il 'Jj eq 25 cm % live stony Code' Type" Exposure" Protection'* (» ' N) jate (m) transects stony corals coral cover (#) (#/IOm) (mean ± sd) 220 o U-5 DO DO DO DO IT) DO o o DO DO DO DO O O On o< NO 'Tj* NO DO DO Tf r^‘ in NO -H -H -H -ti -b -H ■tt -H -H -b -H -H -H -H -H -H -H -H o o o o DO 00 o DO o DO o DO DO o (N rn 00 (N fN Os NO On *_■ NO Tf o NO ON (N (N (N 00 DO DO DO (N (N ro ro fO — (N f— NO 00 NO (N NO On o ON DO ON ON NO rf oo ON r- NO o fO rf DO fN NO rr 00 r- m DO NO O DO NO rr 00 c^ (N Tf — NO Tt 00 (N Os (N f*0 ■rr fO 3 a. a. Q. 3 C-L CL CL CL CL CL CL o o o o o o o o CL o o o CL CL o Cl •o "O f?3 c "O "O •o •o •o “O •o ■p *p ? *p -o c CC3 CS c3 c3 c3 T3 -o ■a x. c3 O cd > •s $ :S :S c5 03 C3 # ■n Ti •n ■n T} ■o > > > Cl “O •o •o "O *o c c c c a c (D b (U 'o c c c c c u- CU xJ o 0) s= a a C/D a a =%! o =« a o cy o a -2 CQ u CQ DO DO DO ly^ 00 CL DD c o s: _op -C _op -C Op s: op x: op a x: op X! o c3 O o .2 "cd o o .2 "o o 1/3 X s s s DO 5 CL kJ DO U. y) C/^ DL DO NO ON o On O r-- 00 ON cn (N NO DO (N m • > 0/ c/5 T5 e -o -o 2 "a CO 0) w -o e -o E ^ Ux (L) C<3 L cq u asey's Reef ast End Reef laying Field aho's Reef elly's Caverns napper Hole CO ighthouse Reel eter's Anchor c/5 o X ^ ^ ^ « 2 8- u II -j o II U C/D Cl, Qi 221 C/5 'O C cd E u (U 4= ■*-> (U >% t-i (U U E .2 E o o s IZl ;-< O o c o cd C _o -4— > 2 ’> > c o oa Oi) c dead »n o p o o in o o o o p o o o O o in O fS O o o c/5 m in in o »n o m o m o m o m in m m m m o in rn 00 so r-' On in id fN so cn id 00 in (N CN (N (N (N (N tN rs »n m m o o o »n o o O o in o o m m o »n in r*^ cn 00 so > o in o rn rn o 00 00 so 00 Os in o^ in Os ON o c ro a> 1> a> 4/ 4> 4/ 4J 4/ 4/ V 4/ 4^ 4/ 4> 4/ 4/ 4> 4> •o c •o c *o c *o c •o c *3 C *o c 1) QC CL X D -j -j -I j _i -j u s CL. % 1> (l> *o 00 (N O m Os m (N SO r- 00 - m Id" «n SO It 00 o u u u u u u u u u u u u u u u u u u E -tt o c JC 4> 9t Cd CJh ?2 c75 4/ E „ 4> 4/ cd 4^ H to Cd m c/5 C c/5 13 i| £ Cd « 'E U -o fi; S 4/ 4/ a H o Q. 3 E o O J= 4/ O. cd > •o c s a Disneylan CO c/5 c3 JZ U No Name Q. H (J -2 oa J c/5 U 3 m u .ts « 3 e = Cd < t; ~ (N ro o r-^ 4/^ in o o in in o o o CN in o ■'S' rn ni in in o in O (N lO — in m m m -H o ■tH -+H in m in in in in lO rn ni 10m) sites, in Little Cayman, Grand Cayman and Cayman Brae. Parameter Setting^ Depth Island' Leeward Windward Leeward ± Protected Protected Shallower Deeper Protected Windward windward ± Windward Windward Diameter, stony corals (cm) LC 43.5 ±26.5 48.5 ±40.0 43.5 42.5 47.5 61.5 ±46.5 41.0 ±26.5 GC 44.0 ±29.0 4L5±3L5 42.5 41.5 41,5 42.5 ±30.5 34.5 ± 13.5 CB 28.0±2L5 35.5±3L5 33.0 ±28.0 3 1.0 ±24.5 Height, stony corals (cm) LC 28.5±2L5 31.5 ±32.0 28.5 26.5 30.5 4L5±38.5 26.5 ±22.0 GC 27.0 ±26.0 22.0 ±21.0 24,0 22.0 22.0 24.0 ±23.0 17.5± 11.5 CB 21. 5± 17.5 23.5 ±20.5 23.5 ±20.5 2L5± 16.5 Diseased stony corals (%) LC 7.5 2.5 7.5 8.0 3.0 8.5 4.5 GC 8.0 2.0 5.5 4.0 3.0 5,0 1.0 CB 4.5 0.5 2.5 1.5 Bleached stony corals (%) LC 2.5 2.0 2.5 3.0 2.5 1.0 3.0 GC 0.5 1.0 2.0 3.0 2.0 1.0 5.0 CB 0.5 0.2 0.2 0.5 Recruitment, stony corals (#/0625m^) LC 0.2 ±0.6 0.2 ±0.6 0.3 0.5 0.3 0.3 ±0.8 0.2 ±0.5 GC 0.3 ±0.6 0.3 ±0.5 0.3 0.3 0.3 0.3 ±0.6 0,3 ±0.6 CB 0.2 ±0.3 0.2 ±0.3 0.2 ±0.3 0.2 ±0.4 Recent partial mortality, stony corals (%) LC 2.5 ±7.5 L0±5.0 3.0 6.5 1.5 2.5 ±8.0 2.0 ±7.5 GC 6.5 ± 15.5 3.0± 11. 6.5 6.0 4.5 5.0± 13.5 3.5 ± 13.5 CB L5±4.5 1.0 ±4.5 L0±3.0 2,0 ±7.5 Old partial mortality, stony corals (%) LC 23.5 ±23.5 19.0 ±24.0 21.5 23,5 19.0 22.5 ±24.0 22.0 ±23,5 GC 23.0 ±24.0 19.0 ±24.0 24.0 24.5 21.5 2L5±24.5 20.5 ±23.5 CB 24.5 ±27.5 15.0 ±23.5 20.5 ±26.0 17.0 ±23.5 Standing dead, stony corals (%) LC 0.5 0.1 0.3 1.0 0.2 0.1 0,3 GC 0.5 2.0 2.0 2.5 2.5 1.5 1.5 CB 3.0 3.0 3.5 2.0 Relative abundance, crustose corallline algae (%) LC 39.0 ±22.0 35.5 ±22.5 39.0 40.0 36.0 34.5 ±25.0 39.0±21.0 GC 36.5 ±20.5 63.5 ±20.5 46.5 51.5 57.5 5L0±2L0 55.5 ± 19.5 CB 30.5 ±22.0 26.5 ±25.0 31.5 ±24.0 21.5 ±22.0 Relative abundance, macroalgae (%) LC 42.0 ±21.0 36.5±2L0 42.0 44.0 37,5 29.0 ±20.0 44.0±2L5 GC 39.5 ±26.0 15. 5± 19.0 39.0 39.0 27.0 27.5 ±22.0 32.0 ±20.5 CB 62.0 ±22.0 56.0 ±26.0 55,5 ±24.5 65.5 ±24.0 Macroalgal Index LC 83.9 93.1 82.2 65.1 89.6 73.6 90.7 GC 55.2 19.8 54.8 54.7 37,2 53.1 23.1 CB 175.0 92.7 100.3 197.3 Relative abundance, turf algae (%) LC 19.5 ± 19.5 27.5 ±22.5 19.0 16.0 26.5 36.5 17.0 ± 18.5 GC 24.0 ±25.0 2L0± 19.0 14.5 9.5 15.5 2L5±2L5 12.0 ± 16.0 CB 7.5 ± 11.5 17,5 ± 19.0 13.0± 16.0 13.0± 15.5 'LC = Little Cayman; GC = Grand Cayman; CB = Cayman Brae ^None of the sites at CB are considered protected leeward with the possible exception of the Bert Brothers location which is a marginal protected leeward location. Only one site on LC would be considered protected windward (Wreck). Table 4. Algal characteristics, density of stony coral recruits and Diadema antillarum (mean ± standard deviation), by site in the Cayman Islands. 224 225 “O 0 01 ■o •D C "O o •u o a. T3 c o k* cu "O C3 "O c *5 II "O c c3 1) (U GC33@ Grand Cayman GC21 GC20 GC30 GC29 GC31 0 5 10 Kilometers Little Cayman - ± Little Cayman: Cayman Brae 100 Kilometers j ■ ; LC03 LC15 LC14 Grand Cayman 4 Kilometers LC09 LC12 LCIO® LC02 @ LC18 LC06 LC04 LC16@ LC01 ® LC11 Figure 1. AGRRA survey sites in Grand Cayman and Little Cayman, Cayman Islands. See Table 1 for site codes. STATUS OF CORAL REEFS OF LITTLE CAYMAN AND GRAND CAYMAN, BRITISH WEST INDIES, IN 1999 (PART 2: FISHES) BY CHRISTY V. PATTENGILL-SEMMENS* and BRICE X. SEMMENS^ ABSTRACT The fish assemblages at 33 sites around the islands of Grand Cayman and Little Cayman were assessed in June 1999 for the Atlantic and Gulf Rapid Reef Assessment initiative using belt transects and Roving Diver Technique surveys. A comprehensive species list, with 58 new records, was compiled for the Cayman Islands based on these data and survey data from the Reef Environmental Education Foundation database. In general, the reefs on Little Cayman appeared to support larger and more individual fishes than those of Grand Cayman. A multidimensional scaling ordination plot showed no clear island pattern but did reveal that the windward or leeward location of each site was an important factor affecting fish community composition. All but two sites followed a pattern of distinct windward and leeward clusters, and these clusters also correlated to macroalgal abundance. The relationship between macroalgal abundance and herbivore density was analyzed and significant correlations were found with surgeonfishes (Acanthuridae) and parrotfishes (Scaridae) using multiple regression. INTRODUCTION Fishes have the potential to provide sensitive indices of reef health. Certain predatory fish species dominate the top of coral reef food webs, hence their density reflects a vast number of human and natural disturbances from habitat alteration to direct exploitation (Ferreira et al., 1998). Similarly, the presence and abundance of herbivorous fishes affect algal composition and cover (Ogden and Lobel, 1978). In response to concerns about the widespread deterioration of reef condition in the Caribbean basin, the Atlantic and Gulf Rapid Reef Assessment (AGRRA) initiative was designed to provide a regional perspective using a standardized methodology. The rapid assessment protocol is focused on three main components of the reef community: stony corals, fish, and algae. As part of this initiative, the reefs of Grand Cayman (GC) and Little Cayman (EC) were assessed in June 1999. ' Reef Environmental Education Foundation, P.O. Box 246, Key Largo, FL 33037. Email: christy@reef org ^ University of Washington, Dept, of Zoology, Box 351800, Seattle, WA 98195-1800. 228 The Cayman Islands are a British Crown Colony located in the western Caribbean. The three islands lie between 19° 15' and 19° 45' N latitude and between 79° 44 ' and 81° 27' W longitude (Fig. 1). GC is the largest and most populous. LC lies approximately 145 km to its east-northeast and is about 10 km from Cayman Brae. The three islands are limestone, horst-and-graben structures associated with the Cayman Ridge (Jones, 1994). Freshwater is scarce and the islands lack rivers and streams. The fringing reefs that surround most of the islands contain shallow reef crests (rubble ramparts) as well as mid-shelf and shelf-edge fore reefs (Blanchon and Jones, 1997). These fringing reefs are particularly well-developed on the windward (eastern and southern) coasts of both islands. Other submerged benthic habitats include seagrass beds and mangrove fringes. The level of human disturbance on GC is significantly greater than on LC, which is relatively remote and undeveloped. Anthropogenic impacts on GC reefs include habitat destruction from anchors and increased suspended sediment load from dredging and mangrove removal. Fishing pressure is considerably greater on GC than around LC. Five spawning aggregations of Nassau grouper {Epinephelus striatus) have been heavily harvested (during the 2002 spawning season, all but one had been depleted). Five hundred local residents are licensed to snorkel with spearguns. Fish pots (Antillean Z- traps) probably represent the biggest threat to the fish communities of both islands (personal observations). An extensive marine park system was established in the Cayman Islands in 1986. Reefs in marine park and replenishment zone areas are protected from fish traps, spearguns, anchoring, and line fishing, although line fishing from shore and beyond the drop-off (shelf edge) is allowed. The Cayman Islands’ Department of the Environment maintains a system of 257 permanent mooring buoys throughout the three islands. The benthos of the Cayman Islands has been well studied, including descriptions of the coral communities, reef status, and analysis of spatial patterns (Roberts, 1988; Logan, 1994; Roberts, 1994). In contrast, apart from descriptions of Nassau grouper spawning aggregations (Colin et al, 1987; Tucker et al, 1993), there are few scientific descriptions of its reef fishes. However, Burgess et al.’s (1994) taxonomic review of collection expeditions contained an annotated list of 381 species known to occur in the Cayman Islands, including the endemic y-lined blenny {Starks ia y-lineata) described by Gilbert (1965). Since 1994, fish sighting and relative abundance data have been collected around the Cayman Islands as part of the Reef Environmental Education Foundation (REEF) Fish Survey Project, an ongoing volunteer monitoring effort. REEF volunteers use the Roving Diver Technique (RDT) (Schmitt and Sullivan, 1996) and the survey data are maintained in a publicly-accessible database. By the end of 2001, the REEF database contained over 40,000 surveys from over 2,000 sites, including approximately 2,200 surveys from the Cayman Islands. This paper describes the fish assemblages of the Cayman Islands using the 1999 AGRRA data for GC and LC, along with REEF data from the two islands collected between 1994 and 2001. An updated species list and comparisons between islands and among sites are provided. The relationship between herbivorous fishes and macroalgal abundance is also investigated. 229 METHODS In June 1999, AGRRA fish and benthos surveys were simultaneously conducted at 15 sites on GC and 18 sites on LC (Fig. 1, Table 1). Sites were chosen by a mixed representative/strategic strategy: 12 were on the windward sides of the islands and 21 were on their leeward sides (the southwest side of GC was underrepresented). Six sites on LC and three on GC were located within marine park or replenishment zone areas. The benthic component is reported by Manfrino et al. (this volume). To assess the fishes, the AGRRA protocol Version 2.1 was used (Appendix One, this volume). At each site, a team of three (occasionally two) divers conducted at least 10 2 m x 30 m belt transects. Counts of serranids (groupers) were restricted to species of Epinephelus and Mycteroperca’, scarids (parrotfishes) and haemulids (grunts) less than 5 cm in length were not tallied. Each diver also conducted a 45-60 minute RDT survey at each site. All fieldwork was undertaken between 9:00 a.m. and 3:00 p.m. Field identifications were based on Humann (1994), Stokes (1980), and Robins et al. (1986). The fish transect data were entered into a custom AGRRA Excel spreadsheet. REEF provided the RDT data in ASCII format. Using the transects as replicates, the average density (#/100 m^) and size (cm) of each species and family were calculated for each site. Analyses were done at the regional (GC versus LC) and site levels, incorporating reef location (windward, leeward) and benthic parameters when appropriate. The average density and size of each species and family were compared between regions using a t-test after testing the data for normality. Due to confounding factors such as differences in use (e.g., recreation, harvest) and hydrographic features, comparisons between protected (marine park) and unprotected sites were not attempted. The site data were used in a hierarchical cluster analysis using Pearson’s similarity index. The similarity matrix was generated using log-transformed density values for each species documented in at least three (10%) of the sites; the other 22 rare species were eliminated (per Grossman et al., 1982). A two-dimensional multidimensional scaling (MDS) ordination plot was also generated using the similarity matrix. The transect data were also used to investigate interactions between the fish assemblages and the benthic community. This preliminary investigation was focused on herbivore/algae interactions. A regression was calculated on the densities of parrotfish and surgeonfish against percent absolute macroalgal abundance in quadrats (hereafter macroalgal abundance) and height at each site. Other coral factors (percent live coral cover, average colony height, percent diseased colonies) and environmental (windward/leeward) parameters were also plotted against each fish family. All values were transformed prior to regression (transformations were log+1 for fish density and algal height and arcsine of the square root for proportions). The RDT survey data provided species lists, frequency of occurrence, and relative abundance estimates. Percent sighting frequency (%SF) for each species was the percentage of all dives in which the species was recorded. An estimate of abundance was calculated as: abundance score = D x %SF, where the density score (D) for each species was a weighted average index based on the frequency of observations in different abundance categories. Density score was calculated as: D= ((nsxl)+(nFx2)+(nMx3)+(nAx4)) / (ns+ np + um + Ha), where ns, np, nM, and nA represented the number of times each abundance category (Single, Few, Many, 230 Abundant) was assigned for a given species. The RDT data were pooled and compared by island using the Wilcoxon Sign Rank test. Only species that were seen in at least 10% of the RDT AGRRA surveys were included in the analysis (103 species), reducing the effect of rare species (Grossman et al., 1982). SYSTAT 7.0 was used for all the analyses. All expert-level REEF data from GC and EC, including the RDT data collected during the AGRRA expedition, were used to compile a species list of reef fishes for the Cayman Islands (REEF, 2001). RESULTS A total of 341 transects (142 - GC; 199 - EC) and 79 RDT surveys (32- GC and 47- EC) documenting 173 species were conducted at 33 reefs (Table 1). The RDT survey data were added to the existing REEF database. The total number of species recorded by REEF experts on the Cayman Islands between 1994 and 2001 was 275 (Appendix A, this paper). When compared with Burgess et al.’s (1994) ichthyofaunal list, the REEF survey data added 58 new species records for a total of 423 reef fishes documented on the Cayman Islands (five freshwater species, 10 deepwater (>300 m) species, and a misidentification {Stegastes mellis) listed by Burgess et al. (1994) were not included in this tally). The 25 most common species, according to %SF in the REEF database, are noted in Table 2. Parrotfish (Scaridae) was the most abundant family recorded during the belt transects (Fig. 2). Average density of snapper (Lutjanidae) on LC was approximately twice that of GC reefs. Size frequency distributions of carnivores (select grouper genera and all snappers) and herbivores [parrotfish >5 cm, surgeonfish (Acanthuridae), and the 30 E 25 o o % □ Little Cayman • Grand Cayman 20 15 1/5 B V Q 10 B 5 J 0 r J z> f O JO ^ # cf Figure 2. Mean fish density (no. individuals/100 + sd) for AGRRA fishes in Grand Cayman and Little Cayman. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. 231 yellowtail damselfish Microspathodon chiysurus] are shown in Figure 3. Approximately 75% of the carnivores were less than 30 cm in length, and 85% of the herbivores were less than 20 cm in length. T-tests on these data showed that the average density and size for most species and families did not differ between islands. However, many species were reported in RDT surveys with greater than average abundance on the LC reefs (Wilcoxon Sign Rank p<0.0005). In particular, the sighting frequencies of six species of large groupers were considerably greater in LC (Table 3; Wilcoxon Sign Rank p<0.05). Exceptions included yellowtail snapper (Ocyurus chiysurus) and sergeant major {Abudefduf saxatilis), two species that become abundant when fed regularly by divers. Fish feeding is much more commonplace on GC reefs (Burgess et al., 1994; personal observations). Site comparisons at the assemblage level showed no clear, intra-island groupings. However, two distinct clusters were obvious in the MDS plot (Fig. 4 A) and, to a lesser extent, in the cluster diagram (Fig. 4B). The only environmental characteristics significantly related to fish density were reef location (windward/leeward) and macroalgal abundance. The windward (high-wave exposure) or leeward/protected windward (low wave-exposure) location of the sites was an important factor in the MDS cluster for all but two of the sites (LC02 and GC30). Leeward sites also had significantly higher macroalgal abundance than windward sites (45% versus 31%, respectively: F-test p<0.001; multiple R = 0.560). A Carnivores B Herbivores 60 50 ■ GC n=292 □ LC n=666 60 □ LC n=3884 50-1 ■GCn=2825 0-5 6-10 11-20 21-30 31-40 >40 0-5 6-10 11-20 21-30 31-40 >40 Size (cm) Size (cm) Figure 3. Size frequency distribution of (A) carnivores (all lutjanids, select serranids) and (B) herbivores (acanthurids, scarids > 5 cm, Microspathodon chrysurus) in GC (Grand Cayman) and LC (Little Cayman). Total number of individuals counted (n) is given. 232 A MDS ordination plot B Hierarchical cluster c o W) c (D E □ -1 OLC11 LC02O LC40 LC150 GC31 O LC14 GC20O GC210 GC290 LC89®C33 '^y§§6oOLC9 LatBiJ fLkSy’ OGC32 m S OGC23 og'K OLC7 OGC25 OGC27 CGC22 -2 0 Dimension-1 Figure 4. (A) MDS ordination plot (left cluster is windward) and (B) hierarchical cluster analysis of AGRRA reef fish transect data in GC (Grand Cayman) and LC (Little Cayman). Surgeonfish density showed an inverse relationship with macroalgal abundance (p<0.01; r2 = 0.209), whereas parrotfish density was positively related to macroalgal abundance (p<0.01; r2 = 0.215) (Fig. 5). Adding macroalgal height to a multiple regression significantly improved the relationships with macroalgal abundance for parrotfish (p<0.001; r^ = 0.413) and surgeonfish (p^O.OOl; r^ = 0.367) densities. A strong inverse relationship between parrotfish and surgeonfish densities was also found (p<0.001;r^ = 0.384). Figure 5. Regression plot between mean parrotfish density (♦) and mean surgeonfish density (□) (no. indiyiduals/lOOm^) and mean absolute macroalgal abundance by site in the Cayman Islands. 233 DISCUSSION The reefs of the Cayman Islands support relatively diverse and abundant fish assemblages. This richness is probably a result of several factors including high local habitat diversity, a significant (34%) area of coastal reserves, and a reef system that is generally in fair condition (Manfrino et al, this volume). However, significant differences were revealed between GC and LC, most likely a result of the greater anthropogenic impacts on GC reefs. Higher harvest pressure on GC was reflected in the lower density and size of large groupers, parrotfishes and snappers (Table 4) and lower sighting frequencies of large groupers (Table 3). Analyses of RDT data indicated that regardless of commercial importance, the average abundance of most fish species was higher on LC, hence other factors, such as coastal development and water pollution, may also adversely impact fish communities on GC. The site-level transect density data correlated most strongly with relative wave exposure (Fig. 4A,B). Macroalgae were significantly less abundant overall on windward (high-wave exposure) sites than on leeward and protected windward (low-wave exposure) sites, where parrotfish were the most abundant fishes in the transects. It is clear, however, that macroalgal abundance does not by itself adequately explain site-level assemblage composition, given that LC sites had only slight differences between wave- exposed and non wave-exposed sites (Manfrino et al., this volume). The correlation between fish communities at sites with similar wave exposure highlights the effect of physical parameters on fish assemblage structure, and should be taken into consideration in fiiture analyses of fish data for the Cayman Islands. In a simple system, one might expect the presence and density of herbivorous species to be negatively correlated with algal abundance and height. In other words, a site with many herbivorous fish would have relatively low algal abundance due to grazing. Our analysis at the site level indicates that this expectation holds true for surgeonfish. However, the inverse is evident in parrotfish. This implies either or both of the following: 1) there is a direct or indirect interaction between parrotfish and surgeonfish, or, more generally 2) the dynamic spatial and temporal characteristics of reef fish confound simple relationships between resource availability and fish abundance. Recent work on stoplight parrotfish (Sparisoma viride) indicates that whereas there are few, if any, direct interactions between surgeonfish and parrotfish, the use of space on the reef by individual fish is complex (territorial behavior, depth partitioning based on social grouping), and varies as a function of social status and intraspecific interactions (van Rooij et al., 1996a; van Rooij et al., 1996b). Clearly, more research is needed to understand the use of space by reef fish if accurate conclusions are to be drawn from relationships between fish abundance and benthic conditions. One of the crucial tasks that scientists face in implementing a “reef health scale” using AGRRA data is to determine exactly what indicators within the collected data track health. An additional challenge lies in assessing how to evaluate and analyze the broad and complementary set of information collected on fishes, stony corals, and algae. Results from this paper and others in this volume will provide valuable insight on these issues. Due to the inherently complex nature of coral reef communities, the manner in which AGRRA data will dictate a scale of reef condition is most certainly also complex. The negative relationship between surgeonfish and parrotfish at the site level is a good 234 example of how community complexity may confound seemingly logical indicators of reef health such as herbivore biomass. Given our results, it is possible that Cayman Island reefs with similar herbivore biomass constituted by predominately different taxa may reflect dramatic differences in benthic conditions. The disparity between grouper abundance between the transect and RDT data and the dramatic increase in species reported in the Cayman Islands that resulted from the RDT surveys (18% based on the published list by Burgess et al., 1994) highlights the importance of using the two complementary visual fish-survey methods. Because certain fish species dominate the top of coral reef food webs, a baseline of fish community composition and richness provides a useful tool for future assessment of reef health, given that a change in reef communities at lower trophic levels will most likely result in changes in the reef fish community composition (Choat, 1991; Jones et al., 1991). Additionally, because fish tend to be the most charismatic group of reef community members, changes in their community are most likely to be noticed and documented. ACKNOWLEDGMENTS The authors wish to thank the Marine Education and Environmental Research Institute (MEERI) and Dr. C. Manfrino for coordinating this immense project. The data collection assistance of K. Hoshino, B. Hansrod, L. Whaylen, and P. Hillenbrand is also greatly appreciated. Funding support was provided by the Gilo Family Foundation, American Airlines, Island Air, Little Cayman Beach Resort, Southern Cross Club, the Reef Environmental Education Foundation, the Cayman Islands Department of Tourism and Tortuga Divers. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. The REEF volunteers who have contributed surveys from the Cayman Islands are also appreciated. REFERENCES Blanchon P., and B. Jones 1997. Hurricane control on shelf-edge reef architecture around Grand Cayman. Sedimentology 44:479-506. Burgess, G.H., S.H. Smith, and E.D. Lane 1994. Fishes of the Cayman Islands. Pp. 199-228. In: M. A. Brunt and J. E. Davies (eds.). The Cayman Islands: Natural History and Biogeography. Dordrecht (Netherlands). Kluwer Academic Publishers. 576 pp. Choat, J.H. 1991. The biology of herbivorous fishes on coral reefs. Pp. 120-153. In: P. F. Sale (ed.). The Ecology of Fishes on Coral Reefs. San Diego, California (USA). Academic Press, Inc. 754 pp. Colin, P.L., D.Y. Shapiro, and D. Weiler 1 987 Aspects of the reproduction of two groupers, Epinephelus guttatus and E. striatus in the West Indes. Bulletin of Marine Science 40: 220-230. 235 Ferreira, C.E L,, A.C. Peret, and R. Coutinho 1998. Seasonal grazing rates and food processing by tropical herbivorous fishes. Journal of Fisheries Biology 53: Suppl. A 222-235. Gilbert, C.R. 1965. Starksia y-lineata, a new clinid fish from Grand Cayman Island, British West Indies. Notulae Naturae, Academy of Natural Sciences, Philadelphia 379. 6 pp. Grossman, G.D., P.B. Moyle, and J.J.O. Whitaker 1 982. Stochasticity in structural and functional characteristics of an Indiana stream fish assemblage: a test of community theory. American Naturalist 120: 423-454. Humann, P. 1994, Reef Fish Identification (2nd ed.). Jacksonville, FL, New World Publications, Inc. 396 pp, Jones, B. 1994. Geology of the Cayman Islands. Pp. 13-40. In: M. A. Brunt and J. E. Davies (eds.). The Cayman Islands: Natural History and Biogeography. Dordrecht (Netherlands). Kluwer Academic Publishers. 576pp. Jones, G.P., D.J. Ferrell, and P.F. Sale 1991. Fish predation and its impact on the invertebrates of coral reefs and adjacent sediments, Pp. 156-178. In: P.F. Sale (ed.). The Ecology of Fishes on Coral Reefs. San Diego, California (USA). Academic Press, Inc. 754 pp. Logan, A. 1994. Reefs and lagoons of Cayman Brae and Little Cayman. Pp. 105-124. In: M.A. Brunt and J.E. Davies (eds.). The Cayman Islands: Natural History and Biogeography. Dordrecht (Netherlands). Kluwer Academic Publishers. 576 pp. Ogden, J.C., and P.S. Lobel 1978. The role of herbivorous fishes and urchins in coral reef communities. Environmental Biology of Fishes 3:49-63. REEF 2001. Reef Environmental Education Foundation. World Wide Web electronic publication, www.reeforg, date of download (31 December 2001). Roberts, H.H. 1988. Grand Cayman. Swamps and shallow marine substrates 1 and 2, Cayman Islands 1; 25,000 (map). U.K.: Overseas Development Natural Resources Institute. Roberts, H.H. 1994. Reefs and lagoons of Grand Cayman. Pp. 75-104. In: M. A. Brunt and J, E. Davies (eds.). The Cayman Islands: Natural History and Biogeography. Dordrecht (Netherlands). Kluwer Academic Publishers. 576 pp. Robins, C.R., G.C. Ray, and J. Douglass 1986. Peterson Field Guides-Atlantic Coast Fishes. New York, NY. Houghton Mifflin. 354 pp. Schmitt, E.F., and K.M. Sullivan 1 996 Analysis of a volunteer method for collecting fish presence and abundance data in the Florida Keys. Bulletin of Marine Science 59:404-416. 236 Stokes, FJ. 1980 Handguide to the Coral Reef Fishes of the Caribbean and Adjacent Tropical Waters Including Florida, Bermuda, and the Bahamas. New York, NY. Lippincott and Crowell, Publishers, 1 60 pp. Tucker, J.W., P.G. Bush, and S.T. Slaybaugh 1993. Reproductive patterns of Cayman Islands Nassau grouper {Epinephelus striatus) populations. Bulletin of Marine Science. 52:961-969. van Rooij, J.M., E.D. Jong, F. Vaandrager, and J.J. Videler 1996a. Resource and habitat sharing by stoplight parrotfish, Sparisoma viride, a Caribbean reef herbivore. Environmental Biology of Fishes 47:81-91. van Rooij, J.M., F.J. Kroon, and J.J. Videler 1996b. The social and mating system of the herbivorous reef fish Sparisoma viride: one-male versus multi-male groups. Environmental Biology of Fishes 47:353- 378. 237 00 TJ G a c cd B >> cd u -o c c3 o T3 C cC C cd B >, u in rf vS +1 -H +1 in in p in (N in (N (N ^ in in in p in vd vd in vd 00 +1 +1 +1 +l +1 o vd m m in in in rf ^ in ^ Os NO OCNCNVOOOnifNeNCN^ 00 ^ ON OV oo ^ »/l W} ^ ^ m — — 00 m o o o r*- o os o 00 Qi o o 0 0 4^00 *o -O *0 .S -o *0 ^ C C c c c c o 'I '% '? ■? o. tt'S'o.D.iS^wSSo. OQ.UO"''''''- Q. O. Q. O. O g « -^s « s ^ Q. § D. O. o J> O, u o ^ M toM g " “ o. a. o. O p " fNinsor***ooosO(Nfnvot^ oooooo^ — — — — UCJUUUUUUUUU 00 »— fo tt — Tf in — ooo — — — U U U U U U U J J nJ J J — +< +1 o o 2 ZZ ^ o\ ^ Os fNvo(nin! .5>C .:^ £Q Cu S )Z s .« _ a Q u Queen's Throne GC25 replen pro wind Hardpan 19 22.818' 81 17.493' 19-Jun-99 12.1 14.5 ±3.0 GC26 replen pro wind S&G 19 21.202' 81 11.746' 19-Jun-99 13.9 12.5+4.0 Bear's Paw GC27 replen pro wind S&G 19 23.854' 8121.617' 20-Jun-99 10.7 15.5 ±3.5 GC19 open wind High S&G 19 19.058' 81 04.484' 17-Jun-99 6.7 23.0 ±9.5 GC20 open wind High S&G 19 20.002' 81 04.596' 17-Jun-99 8.9 18.0 ±4.5 238 £; ^ C o *0 O > « S +1 > « § - p g m fN O O O +1 O si 'S -S li) ^ S' S' s s •SJ Appendix A. Cayman Islands Species List. Data compiled from the REEF database, using expert sightings from 1994 through 2001. A total of 670 expert surveys (32- Cayman Brae; 258- Little Cayman; 380- Grand Cayman) reported 276 species. For each species, percent sighting frequency (%SF) and density score (DEN) are given. Fifty-eight species previously unreported from the Cayman Islands are listed and indicated by an asterisk (*). 241 Appendix A, Continued 242 o P o (N p p p ■of P P p P r-; O p o Tf o o c « o o’ ^ § C rr b OX) U o ^ "O C “ » ^0/23 >- U too U • S Tto ' to) on •X ® on >. XI o a 1/1 c/i c« o •2 2^ .Q c/) Q ^ o s O o o a O X o 13 U Oh X o >1 ”5 o O ■o to) CQ to) .. G- T3 75 « rS CL, O >1 X o 2^ 0.6 >> .^O 2 ^ O "o CL 73, •C ■d o X ^ X >1 o ^ ^ "o ^ ^ o o ;§ -o O o O "O <^1 toi ^ — X § o c .S 5 too .o op is c o — ® c d c3 x “ to) 03 =0 to) t: o to) X 4r .0) Q. o o to> X o >.o o & O S >1 X o >. ■q ”0 "o o a § too to) CL c cto CQ o. o O C2 c/20onU>'ooOoni— !>-' o o B to) on >-> X ^ o X O to) « I ■St' X ^ VI VI — O 3 « O cri H eo >« O S to >«» f a Q c o 2 -» >% JJ C C c c: >-» 3 5 iri c H p2 CQ ^ U S S PL Cl a a^ -a a S N 55 JO "O *-£3 'a c a o “2 "g B s. ® a ■« ^ U a _ a .to to 2 to 2 I Cl a ^ a a a a a a a a a “a "a "a 2 2 2 a a a a a a U U <0 a a to a 2 _a a “a r a « to to a a -2 -2 a a a a a a »2; -a U U 1 I -a S a -a 2 8 a a ®X3 ^ B 2 O tu u i: a a « > 2 g, ^ ® -2 ^ S' w a « a Q Q 0/ a .2 2 'S a B tos to) V? X « a to) Q Q U .LJ to .S .2 ■*- 'C! W Ri 93 a a -a Co ;s a •& “a a 2 a' Q. _ g "a; a. 2 GL bo X -a a o W * a a .top -a a a C! 2 *•*>»* ^ a a -a to" a Ji -a ~c^ 1-f S -2 5 •a a S g a -a tC S -2 ^ ^ 5r to Cl a a a a a -a 03 a S .a to S a I a a 2 5 CQ c« 'O C ’S -S t i h. u * 243 244 245 Scientific Name Common Name SF% DEN Scientific Name Common Name SF% DEN 246 247 Plate 6A. “Recent mortality,” as in this Montastraea annularis lobe, is defined as any non-living parts of the coral in which the corallite structures are white and either still intact or covered by a thin layer of algae or fine mud. (Photo Kenneth W. Marks) Plate 6B. Recent mortality resulting from parrotfish bites, most commonly observed in the Montastraea annularis species complex (as shown) and Colpophyllia natans, is characterized by partial loss of the skeleton along with the overlying living tissues. (Photo Robert S. Steneck) Figure 1. AGRRA survey sites in Cahuita, Costa Rica (modified from Cortes, 1998). A RAPID ASSESSMENT AT CAHUITA NATIONAL PARK, COSTA RICA, 1999 (PART 1: STONY CORALS AND ALGAE) BY ANAC. FONSECA E.‘ ABSTRACT Live stony coral cover, which has deteriorated in Costa Rica's Cahuita National Park, was 2-3% in three reefs during October 1999. About eight percent of all “large” scleractinians (>25 cm in diameter) were diseased. The number of scleractinian species and the densities of recruits and Diadema antillarum were greatest in a carbonate hardground, where total (recent + old) partial-colony mortality of large scleractinians and macroalgal abundance were lowest. Colonies of Acropora palmata that were mostly dead were found only in a shallow patch reef where old partial-colony mortality was highest and where the largest corals were found. Recent partial-colony mortality was lowest in an outer fore-reef habitat. Recovery of deforested watersheds, the main source of the sediment stress in this and other reef systems on the Caribbean coast of Costa Rica, is urgently required. INTRODUCTION Cahuita National Park, situated 35 km south of the city of Limon, has the largest fringing reef on the Caribbean coast of Costa Rica (Cortes, 1998). Its main outer reef, situated 1 km from the coast, arcs around the northern tip of Punta Cahuita, runs southeast about 5 km, and bends inshore towards Puerto Vargas (Fig. 1). A narrow spur- and-groove system reaches depths of 10 m on the outer fore reef. Several small patch reefs occur among meadows of seagrass {Thalassia testudinum, Syhngodium filiforme), algae, and coral fragments in a 3 m-deep lagoon shoreward of the outer reef. Several shallow carbonate hardgrounds (7- 10m) are located south of the main reef and offshore of Puerto Vargas beach. A small (500 m long) inner reef is located 100 m off the north end of the beach. Stony coral cover (measured by the chain transect method) averaged 40% across this inner reef in 1981 (Cortes and Risk, 1984). Rainfall at Cahuita averages about 300 cm/year. The nearshore current, which flows from northwest to southeast, is strong, and wave energy is high, but the tidal amplitude is low (30-50 cm) (Cortes, 1998). Stony coral growth rates and diversity are low. As summarized by Cortes and Risk (1984), species inhabiting the Cahuita reefs are ' Centro de Investigacion en Ciencias del Mar y Limnologia (CIMAR), Ciudad de la Investigacion, Universidad de Costa Rica, San Pedro, P.O. 2060, San Jose, Costa Rica. Email: afonseca@cariari.ucr.ac.cr 250 considered to be good sediment rejecters (e.g., Siderastrea, Diploria), or to have morphologies that inherently facilitate sediment removal (e.g., frondose Agaricia agaricites) and/or efficiently intercept light (e.g., foliacious Porites astreoides). Since 1970, 600 ha of the reef have been afforded partial protection by the creation of the Cahuita National Park (Cortes and Guzman 1985). However, deforestation of the highlands and inappropriate agricultural practices, particularly on the coastal banana plantations, had already increased the input of noncarbonate (terrigenous) sediment and other pollutants in runoff, especially from Rio La Estrella. Physical damage to the reefs from visitors (trampling, diving, and anchoring ) has also increased (Cortes, 1994). Both local inhabitants (who are allowed to fish with hook-and-line) and tourists walk through the shallow lagoon causing further disturbance to shallow habitats. Other stressors have included the 40% mortality of stony corals, particularly Acropora palmata and^. cervicornis, during the 1983 El Nino-Southem Oscillation (ENSO) event and the near-demise of Diadema antillarum in 1983 and again in 1992. Shallow-water colonies of Agaricia agaricites and Porites porites were severely impacted by the 50 cm uplift of the coastline during the 1991 Limon earthquake (Cortes, 1994). By 1993 stony coral cover had fallen to 1 1% in 5 m at the Caribbean Coastal Marine Productivity (CARICOMP) monitoring site on the inner reef near Puerto Vargas (Cortes, 1 994) where macroalgae were reported to be smothering stony corals and possibly affecting larval settlement. More recently, mild bleaching occurred in the reefs at Cahuita in 1998 (Garzon-Ferreira et al., 2000). The purpose of this study was to characterize and update the condition of the coral and benthic algal communities in Cahuita National Park in October 1999 at the end of the millennium using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol. Similar data were obtained the following year south of Cahuita close to the Panamian border in the fringing fore reef off Manzanillo, where live stony coral cover was 1.9% in 1988 (Cortes, 1992). Results of the companion fish censuses are given in Fonseca and Gamboa (this volume). METHODS Three sites (Fig. 1) that are each considered representative of a habitat type in Cahuita National Park were strategically selected for survey by two divers. • Meager Shoal is a circular 1 0,000 m hardground located south of the southern end of the outer reef. Seaward of Puerto Vargas beach, at an average depth of 7.0 m and about 1 km from the coast; it was chosen for its high coral diversity. The sediment that surrounds this bank is predominantly terrigenous muds and easily resuspended by currents. 'y • Eduardo Garden, a shallow (2 m) patch reef of 6,060 m in the lagoon between the outer and inner reefs, is about 200 m from the coast. Constructed of largely dead colonies of Acropora palmata, it is known to be in better condition than other lagoonal sites, and local guides take tourists snorkeling here. Eduardo Garden was the 251 “outer patch” reef in Cortes and Risk’s (1985) investigation of siltation stress at Cahuita. • Chance Mouth is a shallow (~5.5 m) spur-and-groove formation in the area of maximum topographic relief in the outer fringing reef. Located about 1 km from the coast, Chance Mouth was chosen because it is adjacent to the mouth of the only natural boat channel crossing the outer reef crest. The shallow fore reef that was surveyed at similar depths (4 m) off Manzanillo is a low-relief platform. The following modifications were made to the AGRRA Version 2.2 benthos protocol (see Appendix One, this volume). The maximum diameter of every stony coral covered by the transect line was recorded. Colonies of Siderastrea were only identified to genus due to difficulties with species identification in the western Caribbean. Millepora was not included in the surveys of individual corals. Coral sizes were measured to the nearest cm. Sediment was brushed gently off the substratum before estimating the abundance of crustose coralline algae. The abundance of sponges and other invertebrates in the algal quadrats was recorded. We used Humann (1993) for a field guide. Species diversity (H’), equity (J) and Morisita similarity (Mi) indices, based on maximum colony diameters, were estimated for all scleractinians under the transect lines. Coral diversity indices were compared using a Student t-test. RESULTS Stony Corals The dominant “large” scleractinians (>25 cm in diameter) in the transects were Siderastrea, Porites astreoides and Montastraea faveolata in Meager Shoal, Acropora palmata, Siderastrea and Diploria strigosa in Eduardo Garden, and Agaricia agaricites and Siderastrea in Chance Mouth. Only Siderastrea was relatively abundant in all three sites. Species diversity differed significantly among sites (p<0.0001), being highest in Meager Shoal, which had twice as many large scleractinian taxa (eight versus four, respectively) as were present in each of the other two reefs (Table 1). The equity index was slightly higher in Eduardo Garden. The similarity of scleractinian taxa was somewhat greater between Meager Shoal and both Eduardo Garden (Mi=0.48) and Chance Mouth (Mi=0.41) than between Eduardo Garden and Chance Mouth (Mi=0.19). Although the average density of large scleractinian corals in all three reefs was approximately 0.5 colonies/m (Table 1), the live stony coral cover (scleractinians and Millepora) was only 2-3%. The mean size (as maximum diameter and height) of the large corals was substantially greater in Eduardo Garden, where A. palmata and Siderastrea were especially large (Tables 2, 3), than in the other two reefs. Recent partial-colony mortality (hereafter recent mortality), which averaged <10% in most surveyed scleractinians, was highest (mean=3%) at Meager Shoal. Estimates of old partial-colony mortality (hereafter old mortality) were more variable (-10-33%) with 34-78% of the colonies in Eduardo Garden and Meager Shoal, respectively, having less than 10% old mortality. The highest liveidead ratio (mean=7) was found in Meager Shoal. 252 Bleaching was only seen in Meager Shoal, where 2% of the large scleractinians were affected (Table 2). The percentage of large colonies affected by disease varied from 7-9%. White-plague disease affected proportionately more colonies of Siderastrea, P. astreoides, D. sthgosa and A. agaricites in Chance Mouth and Meager Shoal than in Eduardo Garden. Black-band disease, found only in Chance Mouth, had infected one colony of Siderastrea. One colony of Diploria strigosa that appeared to have red-band disease was recorded in Eduardo Garden. The margins of less than five percent of the live scleractinians were being conspicuously overgrown by sediment-trapping turf algae in all reefs, or by macroalgae in Eduardo Garden and Chance Mouth. Recruitment by scleractinians, primarily Siderastrea (12/18 recruits), was highest in Meager Shoal (Table 4). Recruits were rare in the other two reefs: one Siderastrea recruit was seen in Eduardo Garden, whereas two each Agaricia agaricites and Dichocoenia stokesii had recruited in Chance Mouth. The following year at 4 m in the Manzanillo fore reef, where the mean live stony coral cover was 1.5%, large scleractinians averaged 32 cm in diameter (Tables 1, 2). Fifteen percent (7/47) of the surveyed corals were scored as diseased, including two colonies of Siderastrea with circular lesions (“white spots”) of unknown origin. Only one recruit of Porites astreoides was found during the entire survey. Algae and Diadema antillarum Turf algae (and the cyanobacterium Schizothrix) constituted the predominant algal functional group in Meager Shoal (mean relative abundance = 67%) and were somewhat more abundant than crustose coralline algae in Eduardo Garden and Chance Mouth (Table 4). The relative abundance of macroalgae, of which the most common were Dictyopteris, Sargassum and Halimeda, varied from 14-25%. Macroalgal height and macroalgal indices (relative abundance of macroalgae x macroalgal height) were both substantially higher in Chance Mouth (means = 5.5 cm and 134, respectively) than elsewhere. Sponges and other invertebrates were less abundant in Eduardo Garden than in Chance Mouth and Meager Shoal. Tall macroalgae (mean=7.5 cm) predominated at Manzanillo in 2000 where the relative abundance of turf algae and crustose corallines were each less than 25% (Table 2). The mean density of Diadema antillarum at Cahuita in 1999 was considerably greater in Meager Shoal than in the other two reefs (~60 versus 1-7 individuals/ 100m , respectively). Diadema antillarum densities in Manzanillo averaged 10 individuals/ lOOm^ in 2000. DISCUSSION The disparity in live stony coral cover between the 1999 AGRRA transects (<3% in three reefs) and those in the inner reef at the 5 m Caribbean CARICOMP site, which have averaged 1 1% since 1993 (Cortes, 1994; Fonseca, 1999, CARICOMP unpublished report), is due partially to natural differences among the reefs and partially to differing methodologies. AGRRA data are based on projected length under a line and 253 underestimate cover relative to values obtained by counting chain links draped over the contours of the substratum. Live stony coral cover off Cahuita was lowest in Meager Shoal, where large (>25 cm in diameter) scleractinians were the most diverse, smallest in size, and had low partial-colony mortality values (Tables 1, 2, 3) and primarily laminar or crustose morphologies. Ambient illumination on the substratum is low: the muddy sediment surrounding the hardground is easily resuspended by currents and it is slightly deeper (7 m versus 5 m, respectively) than the mean annual vertical Secchi depth in this location (Fonseca, 1999, CARICOMP unpublished report). The bleached scleractinians found here in 1 999 probably represented colonies that, due to the chronically low-light levels, had not yet recovered from the mild bleaching event the previous year. The slightly higher values of live coral cover (Table 1) as well as the more robust morphologies of the dominant scleractinians in Eduardo Garden and Chance Mouth may be related to their shallower and more highly illuminated locations. Eduardo Garden and Chance Mouth are surrounded by sandy, autochthonous sediment which presumably also contains fewer pollutants (see Cortes, 1994) than the riverine muds that are more likely to accumulate in Meager Shoal. The high percentage of colonies with signs of active disease (7-9% at Cahuita in 1 999; 1 5% at Manzanillo in 2000), coupled with the low cover of live stony corals in these Caribbean reefs, is of great concern. Estimates of recent partial mortality for large scleractinians were low in both years (0.5-3%). However, as exposed skeletal surfaces are rapidly covered with sediment and/or overgrown by algae (personal observations), snapshot assessments of recent mortality, as employed in the AGRRA benthos protocol, are likely to underestimate the magnitude of recent soft tissue loss in Costa Rica's coastal scleractinians. Old (and total) partial-mortality values were also fairly low (<20%) in all but Eduardo Garden, which had a higher proportion of large scleractinians {Acropora palmata and Diploria strigosa) that were mostly or entirely dead. Given its shallow location in the lagoon, some of these corals may have died during the 1983 ENSO event; alternatively, the A. palmata may have been infected by white-band disease, as seen elsewhere throughout the Caribbean (Aronson and Precht, 2001). Densities of Diadema antillarum are still much lower than were found in Cahuita during the late 1970’s and early 1980’s (Cortes, 1994). Nevertheless, individuals were reasonably common in Meager Shoal (Table 4), where several Diadema were seen spawning and where they appeared to exert more control on benthic macroalgae than in the two other surveyed reefs. However, it is also likely that benthic algal productivity in this hardground is limited by the dim ambient illumination. Crustose coralline algae and macroalgae were relatively more abundant in Eduardo Garden and Chance Mouth, where Diadema was comparatively rare (Table 4). Less fishing occurs in these two reefs, yet their larger populations of herbivorous fishes (Fonseca and Gamboa, this volume) appear less effective at consuming macroalgae (Hughes et ah, 1987; Hughes, 1994) than the Diadema in Meager Shoal. In Chance Mouth, where the higher surge and wave action may inhibit feeding by herbivorous fishes (Hay, 1981), the macroalgae were taller and the macroalgal index (a proxy for biomass) was correspondingly higher. Diadema were also scarce and reef fishes were far less abundant in Manzanillo in 2000 than in Chance Mouth in 1999 (Fonseca and Gamboa, this volume). Macroalgae were more abundant and 254 larger here than had been found the previous year in Cahuita. Hence coral reefs that are currently in poor condition are not restricted to Cahuita National Park. Between 1 993 and 1 999, suspended particulate matter increased from 9 mg/1 to about 20 mg/1 at monitoring sites in the Cahuita National Park, including Meager Shoal and Eduardo Garden (Cortes, 1994; Fonseca, 1999, CARICOMP unpublished report), which doubtless is contributing to the continued low stony coral cover and low diversity of this reef system. Watershed owners must be encouraged to conscientiously and consistently follow the national regulation which requires a 50 m wide forested buffer along the rivers to ensure good water quality and improve conditions in coastal aquatic ecosystems. Moreover, as Cortes and Murillo (1985) have noted, “Coherent units of reefs and terrestrial environs must be considered when establishing a marine park or reserve. Watershed of rivers near the reef must be included in the protected area.” ACKNOWLEDGMENTS 1 would like to thank the Akumal Workshop presenters for training in the AGRRA protocols. The AGRRA Organizing Committee facilitated my participation in the Akumal and Miami workshops as described in the Forward to this volume. 1 really appreciate the help of Justo Lopez, Marta Coll and Odalisca Breedy in the field. I also thank the Centro de Investigacion en Ciencias del Mar y Limnologia (CIMAR) for providing me with field equipment. And finally, I thank Dr. Jorge Cortes for delegating this task to me, for some funding from his own projects (US-AID-CDR, Project TA MOU-97-C14015) and for reviewing the manuscript. REFERENCES Aronson, R.B., and W.F. Precht 2001. Evolutionary paleoecology of Caribbean coral reefs. Pp. 171-233. In: W.D. Allmon and D.J. Bottjer (eds.). Evolutionary Paleoecology: The Ecological Context of Macroevolutionary Change, Columbia University Press, New York. Cortes, J. 1992. Los arrecifes coralinos del Refugio Nacional de Vida Silvestre Gandoca- Manzanillo, Limon, Costa Rica. Revista de Biologla Tropical 40:325-333. Cortes, J. 1994. A reef under siltation stress: a decade of degradation. Pp. 240-246. In: R.N. Ginsburg (compiler) Global Aspects of Coral Reefs - Health, Hazards, and History, 1993. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Cortes, J. 1998. Cahuita and Laguna Gandoca, Costa Rica. Pp. 107-1 13 In: Kjerfve, B. (ed.), CARICOMP. Caribbean coral reef, seagrass and mangrove sites. Coastal region and small island papers, 3, UNESCO, Paris. 255 Cortes, J., and M.M. Murillo 1985. Reefs under mud. In: Symposium on Endangered Marine Animals and Marine Parks, Paper 50. Marine parks, Cochin, India 1 :487-490. Cortes, J., and M.J. Risk 1984. El arrecife coralino del Parque Nacional Cahuita, Costa Rica. Revista de Biologia Tropical 32:109-121. Garzon-Ferreira, J., J. Cortes, A. Croquer, H. Guzman, Z. Leao, and A. Rodriguez- Ramirez 2000. Status of coral reefs in southern tropical America: Brazil, Colombia, Costa Rica, Panama and Venezuela. Pp: 331-348. In: C. Wilkinson (ed.), Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science. Cape Ferguson, Queensland and Dampier, Western Australia. Hay, M.E. 1981. Herbivory, algal distribution, and the maintenance of between-habitat diversity on a tropical fringing reef. American Naturalist 1 1 8:520-540. Hughes, T.P. 1994. Catastrophes, phase shifts, and large-scale degradation of a Caribbean Coral Reef Science 265:1547-1551. Hughes, T.P., D.C. Reed, and M.J. Boyle 1986. Herbivory on coral reefs: community structure following mass mortalities of sea urchins. Journal of Experimental Marine Biology and Ecology 1 13:39-59. Humann, P. 1993. Reef coral identification: Florida, Caribbean, Bahamas. New World Publications, Inc., Jacksonville, FL, 239 pp. Table 1. Site information for AGRRA stony coral and algal surveys in Cahuita and Manzanillo, Costa Rica. 256 z C o X5 (U X ^ c .3> U a> 3 CL cr C/) 0) o c c T3 C C3 CJ3 C o a. 257 C3 O (2 C3 ■t— > c/J O U CO CO U c 0) 1/3 'o (U T3 "O cO -^3 c CO +1 c cO (U D "S s CO X) cO H 1/3 (U +-> X) c _o '*-» 2 ’> (U T3 'O cO T3 C 173 +1 c CO 25 cm diameter) stony corals were fairly abundant (7.5-10.5/10 m), but in two reefs had incurred moderately high values of recent partial-colony mortality (~10%). The major stressors on these stony corals were damselfish bites, damselfish algal gardens, diseases, and a relatively high abundance of fleshy macroalgae. Damselfish densities were probably elevated because their predators have been overharvested. Enforcement of existing fishing regulations and other management actions are necessary to preserve the ecological and touristic value of this reef system. INTRODUCTION The Maria la Gorda coral reef, considered one of the most beautiful, diverse and well-preserved of Cuban reefs, is located in a 1,216 km Biosphere Reserve near the southeastern tip of Bahia (Ensenada) de Corrientes in the Peninsula de Guanahacabibes (Fig. 1). The adjoining land is a karst plain covered by a naturally semideciduous, mesophytic forest that has been little disturbed although it is subjected to a restricted forestry. Annual rainfall averages 100-120 cm and mean air temperatures are 27-28° C in July and 20-22° C in January (IGEO-ICGC-ACC, 1989). Low rocky cliffs with coastal vegetation alternate with small sandy beaches along the coastline. No rivers empty into the bay. * Institute de Oceanologia, Ministerio de Ciancia, Technologia y Medio Ambiente, Ave U No. 18406, Reparto Flores, Playa, Ciudad de La Habana, CP 12100, Cuba. Email: alcolado@ama.cu and alcolado@oceano.inf cu 270 The only human settlement at Maria la Gorda is a small (29-room) dive resort with 50 associated dive sites. La Bajada, a tiny village located about 14 kilometers further north, has 22 houses, a frontier guard post, the Reserve’s Visitors Center, and a meteorological radar station. Its human population totals less than 1 50 inhabitants. The Guanahacabibes peninsula protects Maria la Gorda’s leeward fringing reefs from the prevailing northeastern trade winds and from northerly cold fronts in winter. An irregular series of elongated low-relief (1-2 m) coral lobes (some forming spurs and grooves), each over -100-200 m in length, about 50-100 m wide and oriented parallel to the shoreline, alternate with tongues of sand at depths of about 5 to 9- 1 2 m on a gently sloping reef terrace. Small patch reefs border the lobes on their shallow and deep margins. However, the diving sites preferred by tourists are located in excess of 1 5 m on the reef slope where the reefs are aesthetically more beautiful, biologically more diverse, and have higher topographic relief. Benthic reef condition was first assessed at Maria la Gorda within the framework of the General assessment of the ecological state of the Cuban reefs and monitoring of the regional Cuban CARICOMP Station and the Atlantic and Gulf Rapid Reef Assessment (AGRRA) initiative in July 1999. METHODS The benthic surveys were made by five divers in four reef lobes (one having well- developed spurs and grooves) that were considered representative of the reef terrace habitats in this strategically chosen tourist location. There were two “shallower” reefs in 5-6 m, and two “deeper” reefs that were located in 8 m and in 1 1 m, respectively. Version 2.2 of the AGRRA benthos protocol (see Appendix One, this volume) was followed. We measured the sizes of “large” (>25 cm diameter) stony corals (scleractinains and Millepora spp.) to the nearest 10 cm. Favia fragum and other species that are small as adults were not included in the counts of coral recruits. Sediment was removed from the algal quadrats before estimating the abundance of crustose coralline algae. Practice and consistency training occurred in reefs near Havana before the surveys were initiated. Our guide for coral identifications in the field was Humann (1993). RESULTS Stony Corals Total live stony coral cover was somewhat higher in the two shallower reefs (20.5%, 23.5%) than in the two deeper reefs (15.5%. 17%), as was the density of large (>25 cm diameter) corals (8.5 and 10.5/10 m versus 7.5/10 m ,respectively; Table 1). In all four reefs, the large stony corals were dominated by Montastraea annularis, M. franksi, Siderastrea siderea and M. faveolata (Fig. 2A, Bj. The mean diameter of the large corals was about 50% greater in the two more southerly reefs (one each being shallower and deeper) than in the two more northerly reefs (60.5 and 63.5 cm versus 44 271 Diploria Acropora A n=236 labyrinthiformis cervicornis 3% 2% Diploria strigosa 2% Colpophyllia natans 1% Agaricia ■■ agaricites 7% Montastraea faveolata 15% Montastraea franksi 17% B n=212 Diploria strigosa 2% Colpophyllia - — natans 1% Agaricia agaricites 6% Montastraea ■ faveolata 12% Montastraea franksi 25% Montastraea cavernosa 5% Diploria Acropora labyrinthiformis cervicornis 3% 2% Siderastrea siderea 17% Siderastrea radians 0.4% Porites furcata 0.4% Porites porites 3% Porites astreoides 4% Montastraea annularis 24% Siderastraea siderea 17% - Porites porites 1% - Porites astreoides 1% Montastraea annularis 24% Montastraea cavernosa 6% Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>25 cm diameter) at (A) 5-8 m, (B) 9-13 m in Maria la Gorda, Cuba. and 46.5 cm, respectively). Overall, M. faveolata formed the largest colonies (to >300 cm in diameter) with no size class predominating, whereas most colonies of M. franksi and M. annularis were <70 cm in diameter, while few S. siderea exceeded 50 cm in diameter (Fig. 3). Large stony corals in the two southerly reefs also had somewhat higher percentages of each of the following relative to the northerly reefs: recent partial-colony mortality (9 and 10.5% versus 3 and 5%); old partial-colony mortality (28.5 and 34.5% versus 24.5 and 27%); “standing dead” corals having 100% mortality on their upper 272 Figure 3. Size-frequency distribution in cm of colonies (>25 cm diameter) of Montastraea annularis, M. franks i, M. faveolata and Siderastrea siderea at 5-13 m in Maria la Gorda, Cuba. Figure 4. Number of colonies with old partial colony mortality and recent partial colony mortality for all stony corals >25 cm diameter at 5-13 m, in Maria la Gorda, Cuba. colony surfaces and still in growth position (2 and 4.5% versus 1 and 1.5%); The percentage of stony corals that were bleached was also higher in the two southerly reefs (3.5 and 5.5% versus 1 and 2%, respectively; Table 2). Nevertheless, the majority of these colonies had no recently dead tissues (Fig. 4). 273 An average of 3-6.5% of the large stony corals in each reef showed signs of disease. Most commonly noted were white-plague disease in M annularis, M. faveolata and M franksi, black-band disease in Diploria strigosa, white-band disease in Acropora cervicornis, and dark spots disease in S. siderea. In addition, some unusually darkly pigmented tissues were seen in Agaricia agaricites (two colonies) and Colpophyllia natans (one colony). The species that were either pale or partially bleached were S. siderea, M. annularis, M. franksi, M. faveolata and P. astreoides. Colonies of S. siderea, M. faveolata, M. franksi and M annularis were found with deep parrotfish bites, while damselfish and their algal gardens were particularly common in D. strigosa and D. labyrinthiformis. The density of stony coral recruits varied from 0.1-0,2/0625m^ (Table 3) which is equivalent to 1.6-3. 2 /m^. Siderastrea siderea, Porites porites and P. astreoides dominated this assemblage (Fig. 5). n=34 Eusmilia . fastigiata 3% \ Manicina areolata 6% Madracis decactis 6% Stephanocoenia intercepta 9% Dichocoenia stokesi 3% Agaricia humilis 9% Unknown 3% Siderastrea siderea 20% " Siderastrea radians 3% Agaricia agaricites 3% f ' Montastraea faveolata 3% Porites porites 15% Porites astreoides Montastraea annularis 3% Figure 5. Species composition and mean relative abundance of all stony coral recruits (<2 cm diameter) at 5-13 m, in Maria la Gorda, Cuba. Algae and Diadema antillarum In terms of relative abundance, macroalgae were the predominant algal functional group (46-60%) in all four reefs, with crustose coralline algae being the least abundant (13-18%; Table 3). Macroalgae were dominated by Dictyota divaricata, Lobophora variegata and Sargassum hystrix whereas green calcareous algae were scarce. Mean macroalgal heights were 2-2.5 cm, and macroalgal indices (relative abundance macroalgae x mean macroalgal height) ranged from 97 to 150. Densities of Diadema antillarum varied from zero to 4.3/100 m^ (Table 3). 274 DISCUSSION Maria La Gorda is the first Cuban coral reef to have been assessed by the AGRRA protocols. The prevalence of the Montastraea annularis species complex in all four reefs is an indication that this reef is generally in good condition. The relatively limited abundance of the sediment-tolerant Siderastrea siderea (Fig. 2) is suggestive that sedimentation is a less severe stress than in many other Cuban reefs e.g., at Archipelagos Sabana-Camaguey, Cayo Diego Perez, Cayo Cantiles and Cayo Juan Garcia (Alcolado in prep.), Cayo Largo (Alcolado et ah, 2001), and Ciudad Habana (Herrera-Morena and Alcolado, 1983; Alcolado and Herrera-Moreno, 1987). Siderastrea siderea was, however, more common in the deeper reefs where sediment deposition rates may be somewhat higher due to reduced wave energy. Although relatively few stony corals were still affected at the time of our survey in July 1999, the populations at Maria la Gorda had undergone severe bleaching during the 1997-98 El Nino warming event (C. Carrodeguas and N. Capetillo, personal communication). The relatively high estimates of recent partial-colony mortality found in the two more southerly reefs (Table 2) may, in part, represent post bleaching-related mortality and, in part, the effects of diseases that were on-going in summer 1999. Overall estimates of recent partial-colony mortality (3-10.5%) at Maria la Gorda varied from low to moderately high by comparison with results obtained in earlier AGRRA surveys (Kramer and Kramer, 1999; Lang, 1999; Leao et ah, 1999; Steneck and Lang, 1999). Similarly, the mean percentages of old mortality (<35%) can be considered moderate. Damselfish bites (primarily due to Stegastes planifrons) and their algal gardens, a suite of coral diseases, and fleshy macroalgae were the most common stressors of the large stony corals at Maria La Gorda. Injuries inflicted by the damselfish could also have facilitated the spread of coral diseases. The positive relationship between colony diameter and impacts (partial-colony mortality, percent of stony corals with disease and damselfish algal gardens) at the two southerly reefs suggests a gradual process of accumulation of partial-colony death as size (and, to some extent, age) increases. The somewhat elevated densities of S. planifrons presumably resulted from a local deficit of carnivorous fishes (Claro and Cantelar Ramos, this volume). The high relative abundance of fleshy macroalgae (and perhaps the scarcity of calcareous macroalgae. Hay, 1 997) is probably a response to the continued scarcity of the key herbivore, Diadema antillarum (Table 3), and to low herbivory by fishes (Claro and Cantelar Ramos, this volume). Hence, to sustain the Maria la Gorda coral reefs, the following management actions will be necessary: • Avoid the construction of hotels and other major tourism infrastructure within the coastal zone of the reserve. • Strictly prohibit all illegal fishing in the Peninsula de Guanahacabibes Biosphere Reserve, allowing stocks of reef fishes to recover. (Partial enforcement of fishery regulations is ineffective, Watson et ah, 1997.) 275 • Avoid nutrification (Alcolado et al., Institute de Oceanologia unpublished report) of the coastal waters. Currently the resort’s wastewaters are poured into excavations in the karst and probably leak into the sea via groundwater. Thus, an alternate disposal system will be required. • Deploy permanent morring bouys at all dive sites. • Ensure the resort divers are educated in safe reef-diving etiquette, and that the carrying capacity of the reefs (Hawkins and Roberts, 1997) is not exceeded. ACKNOWLEDGMENTS The AGRRA organizing Committee, which facilitated financial support as described in the Forward to this volume, is also thanked for the invitation to collaborate in this regional initiative. We acknowledge the invaluable collaboration with, and support from. Reef Relief. REFERENCES Alcolado, P M., and A. Herrera-Moreno 1987. Efectos de la contaminacion sobre las comunidades de esponjas en el Litoral de La Habana, Cuba. Reporte de Invesigaciones del Instituto de Oceanologia, Academia de Ciencias de Cuba, 68:1-17. Alcolado, P. M., R. Claro-Madruga, B. Martinez-Daranas, G. Menendez-Macia, P. Garcia-Parrado, K. Cantelar, M. Hernandez, and R. del Valle 2001 . Evaluzacion ecologica de los arrecifes coralinos del oeste de Cayo Largo del Sur, Cuba: 1998-1999. Boletln de Investigaciones Marinas y Costeras 30: 109-132. Hawkins, J.P., and C. Roberts 1997. Estimating the carrying capacity of coral reefs for SCUBA diving. Proceedings of the Eighth International Coral Reef Symposium, Panama 11:1923-1996. Hay, M.E. 1997. Calcified seaweeds on coral reefs: complex defenses, trophic relationships, and value as habitat. Proceedings of the Eighth International Coral Reef Symposium, Panama 1:713-718. Herrera-Moreno, A., and P.M. Alcolado 1983. Efecto de la contaminacion sobre las comunidades de gorgonaceos al Oeste de la Bahia de la Habana. Ciencias Biologicas 10:69-86. Humann, P. 1993. Reef Coral Identification. New World Publications, Inc., Jacksonville, FL, 252 pp. 276 IGEO-ICGC-ACC 1989. Nuevo Atlas Nacional de Cuba. Institute de Geografia, Institute Cubane de Geedesia y Cartegrafia y Academia de Ciencias de Cuba. Ciudad de La Habana. 226 unnumbered pages. Kramer, P.A., and P.L. Kramer 1999. Determining spatial pattern in reef cenditien: Applicatien ef AGRRA-RAP te the Andres Island reef system, Bahamas. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration: 121 (Abstract). Lang, J.C. 1999. Rapid assessment at the Flewer Gardens: remarkable reefs fleurishing in the nerthwestem Gulf ef Mexice. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration:\22 (Abstract). Leae, Z., R. Kikuchi, V, Testa, M. Telles, J. Pereira, L. Dutra, and C. Sampaie 1999. First ceral reef assessment in the seuthem hemisphere applying the AGRRA Rapid Pretecel (Caramuanas reef, Bahia, Brazil). International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration: 122-123 (Abstract). Steneck, R.S., and J.C. Lang 1999. Applicatien ef the Atlantic and Gulf Reef Assessment Pretecels aleng Mexice’s Caribbean Ceast 1997-1999: Pre and pest bleaching impacts. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring and Restoration: 1 84 (Abstract). Watsen, M., R.F.G. Ormend, and L. Heliday 1997. The rele ef Kenya’s marine prelected areas in artisanal fisheries management. Proceedings of the Eighth International Coral Reef Symposium Panama 2:1950-1950. 277 cd U cd T3 u O a rS cd S (U C/5 C3 H S- (U 15 s: JZ on o a i2 v2 o cd X) u cd^ T3 ;-c O O cd 'C cd s .s (U >» X a> t> S .2 S CJ in (U T3 cd C B C/3 +1 c cd X5 -o c cd 2 o (U Id i-( o o c o (U o c cd "O C 3 X cd C _o '+-> 2 ’> D TS ^3 cd ’X C cd +1 C cd (U c/3 o t: - < La Misteriosa 50 47 ±24 35 ± 22 18± 17 2.4 ± 3.4 113 0-16 ^Macroalgal index = relative macroalgal abundance x macroalgal height 278 Figure 1. AGRRA survey sites in Maria la Gorda, Cuba. See Table 1 for site codes. RAPID ASSESSMENT OF CORAL COMMUNITIES OF MARIA LA GORDA, SOUTHEAST ENSENADA DE CORRIENTES, CUBA (PART 2: REEF FISHES) BY RODOLFO CLARO’ and KAREL CANTELAR RAMOS* ABSTRACT An assessment of fish community structure was carried out in the fringing reef at Maria la Gorda in the Guanacahabibes Biosphere Reserve, following the Atlantic and Gulf Rapid Reef Assessment protocols. Quantitative “all species” surveys were also taken for comparison with existing information elsewhere. Unexpectedly for a no-take reserve, fish density and biomass were very low in comparison with other Cuban reefs. In the two shallower (5-8 m) reefs, parrotfishes and grunts were numerically dominant, and grunts were dominant in terms of biomass, but most were small-sized species. A balistid (Melichthys niger) dominated one of the deeper (9-10 m) reefs. Groupers and most snappers were uncommon at all depths. The scarcity of medium- and large-sized fishes is a consequence of illegal fishing in the reserve. Reduced predation by piscivores may also have caused the proliferation of damselfishes, some of which are killing corals and may be facilitating the spread of coral diseases. INTRODUCTION The leeward fringing coral reef at Maria la Gorda, a small resort in the Peninsula de Guanahacabibes Biosphere Reserve in western Cuba, is considered one of the nation’s most beautiful and best conserved. The only tourist facility, which has 29 rooms and belongs to a scuba diving resort, is located near the southeastern tip of Ensenada de Corrientes (Fig.l). Not far from this resort is a small settlement with 22 houses, a frontier guard-post, a visitor’s center for the reserve, and a meteorological radar station. There are fewer than 200 inhabitants (including tourists) during the high tourism season (about six months of the year). The Reserve has been a no-take area for fishing since 1 996 by a Resolution of the Ministry of Fishing Industry. Hence its coral reefs would be expected to have a relatively intact fish population. However, there is no existing quantitative information regarding the fish communities in this area. 'institute de Oceanologia, Ministerio de Ciencia, Tecnologia y Medio Ambiente, Ave U No. 18406. Reparto Flores, Playa, Ciudad de La Habana, CP 12100, Cuba. Email: rclaro@oceano.infcu 280 Reef lobes of irregular shapes, but elongated parallel to the bay’s coastline, rise 1-2 m above a gently sloping sandy terrace between depths of 5 and 15 m. Live stony coral cover and the density of “large” (>25 cm diameter) colonies were both considered moderate in July 1999, slightly higher at 5-6 m in two shallow lobes (5-6 m) than in two deeper lobes (8-1 1 m), and dominated by Montastraea annularis, M. franksi, Siderastrea siderea and M. faveolata (Alcolado et al., this volume). Nevertheless, the preferred dive sites for tourists are on the higher relief, structurally more complex, and biologically more diverse reefs that are located at depths of 20-30 m along the outer margin of the terrace. The topography on the reef lobes is very irregular, potentially providing plenty of shelter for fishes and mobile invertebrates. Fishermen who have worked for many years in the area and local villagers emphasized that, until about 20 years ago, the fringing reef supported numerous, large- and medium- sized predatory fishes such as snappers (lutjanids), groupers (serranids), and large-sized parrotfishes (scarids). The Maria la Gorda area was traditionally fished by commercial fishermen using such artisanal fishing gear as fish traps and hooks and lines. Eventually, non-commercial fishers began to use spearguns. Illegal speargun and hook-and- line fishing increased during the early 1990s due to the economic crisis in the country. Small-scale commercial fishing currently occurs outside the reserve, where a few tourists are probably also allowed to fish. The latter is not significant, but due to the small size of the reserve, resident fishes can be harvested close to its boundaries during their local migrations. Illegal subsistence fishing, practiced by local residents and outsiders, continues within and outside the reserve, due to a low level of enforcement of the regulations. Located approximately eight km southwest of the reserve’s boundary, at a depth of 25-40 m in Cabo Corrientes, is a well-known spawning site for mutton snapper {Lutjanus analis) in June, cubera {L. cyanopterus) and dog snapper {L. jocu) in July- August, Nassau grouper (Epinephelus striatus) in December- January, plus black grouper {Mycteroperca bonaci) and yellowfm grouper (M venenosa) in January-March. These aggregations are fished mainly with hooks and lines and fish traps (Claro and Lindeman, in press). The objective of this paper is to contribute to the assessment of reef condition at Maria la Gorda by examining the status of the fish community and its relationship to impacts. METHODS The reef at Maria la Gorda was strategically chosen on the basis of its reputation among divers. Four fish surveys were carried out at two depths (5-8 m and 9-13 m) on representative fore-reef lobes in July 1999 (Table 1). Fishes were visually censused using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) Version 2.2 protocols (see Appendix One, this volume), modified as described below. Quantitative data for the AGRRA fishes in each reef were collected along six belt transects, each 50-m long x 2 m wide, provided quantitative data for the AGGRA fishes in each reef The transect tape was 281 set on the substratum by the diver, and the counts started five minutes later. Counts of serranids were restricted to species of Epinephelus and Mycteroperca; scarids and haemulids (grunts) less than 5 cm in length were not tallied. All surveys were made between 10:00 and 17:00 hours by two divers. For identifications, we used Humann’s (1994) guide to reef fishes. Leather) ackets were classified according to Eschmeyer’s (1998) revision, in which Batistes and Melichthys are in the Balistidae, while Aluterus and Cantherhines are assigned to the Monacanthidae. Biomass estimates of the AGRRA fishes were made using the length-weight relationships given by Garcia- Arteaga et al, (1997) for Cuban fishes. A ranking index (% sighting frequency x % abundance) was calculated for the 25 most abundant of these fishes (data summed over all sites). Species diversity (H’), richness (Ri), and evenness (J) indices for all fish species were estimated on the basis of two qualitative, 30-minute roving diver surveys in each reef We made six additional “all species” belt transects (each 50 m long x 2 m wide) to count and estimate sizes of all fishes in three of the reefs, to compare with similar surveys made at 15 m depth in 16 sites of the Archipelago los Canarreos (SW Cuba) in 1988-1989. Even though the data are not fully comparable statistically, due to the small sample size at Maria la Gorda, the comparison may serve as a primary reference for assessing current conditions in Maria la Gorda’ s population of reef fishes. RESULTS Fish Diversity A total of 88 fish species (Table 2) were found during the eight roving diver censuses (mean = 58/reef, sd= 6). Relative to the two shallow reefs, estimates of species diversity and evenness in the two deeper reefs were slightly lower, while species richness estimates were slightly higher. The most frequently sighted and abundant of these species were the blue chromis (Chromis cyanea), bluehead wrasse (Thalassoma bifasciatum), and bicolor damselfish (Stegastes partitas) (Table 3). In the “all species” belt transects, there were 42-54 species per reef (for a total of 7 1 species). Between 25 and 29 of the AGRRA species were seen in the AGRRA belt transects in each reef (total =45). Highest in the AGRRA hierarchy index were French grunt (Haemulon ftavolineatum), striped parrotfish (Scarus iserti, =S. croicensis, see Eschmeyer, 1998) and blue tang (Acanthurus coeruleus) (Table 4). Fish Density The total density of the AGRRA fishes (Table 5) was nearly identical in the two shallower reefs (-30.3 and 31.2 individuals/ 100 m^) but more variable in the two deeper reefs (-20.3 and 37.5 individuals/100 m^). Herbivorous fishes, primarily small-sized parrotfishes, comprised over half (-52%) of the total numerical count while grunts constituted the second most abundant family overall (Fig. 2). Striped parrotfish dominated in one of the shallow reefs (Sh-1), whereas French grunts were predominant in 282 the other (Sh-2). A large group of black durgeon (Melichthys niger) and blue tang {Acanthurus coeruleus) predominated in one of the deeper reefs (D-1), where yellowtail snapper {Ocyunis chrysurus) were also moderately common. The second deeper reef (D- 2) had the fewest fishes overall and lacked dominance by any species. Surgeonfishes (acanthurids) were slightly more abundant in the two deeper reefs than in the shallower reefs. Densities of snappers, angelfishes (pomacanthids), butterflyfishes (chaeotodontids), and select groupers {Epinephelm and Mycteroperca) each totaled <6 individuals/ 100 m^ for the four reefs. However, a larger number of snappers (O. chrysurus and Lutjanus apodus) were observed close to the edge of the reef (and outside the area covered by the belt transects) during the roving diver censuses at one of the deeper reefs (D-1). Figure 2. Mean fish density (no. individuals/100 ± sd) for AGRRA fishes in Maria la Gorda, Cuba. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. The mean density of fishes in the “all species” belt transects was nearly 20 times greater than that of the selected species in the AGRRA fish transects (541 versus ~30 individuals/100 m^). Damselfishes (pomacentrids) were the most abundant family with a 'y mean density of 126 individuals/ 100 m , and most (>93%) of these were bicolor damselfish (Table 6). At approximately 36 individuals/100 m^, the average density of herbivores (all parrotfishes and surgeonfishes, Microspathodon chrysurus) was almost 20 times greater than the average density of carnivores (all snappers, Epinephelus, Mycteroperca), of which there were only 2.6 individuals/100 m^. 283 Figure 3. Size frequency distribution (in cm) of herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) in Maria la Gorda, Cuba. Figure 4. Size frequency distribution in cm of carnivores, as (A) all lutjanids and haemulids >5 cm, select serranids and (B) all lutjanids and select serranids, in Maria la Gorda, Cuba. 284 Fish Size Most of the key herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) were less than 20 cm in total length (Fig. 3). Large-sized parrotfishes (such as Scarus gnacamaia and S. coelestinus) were absent from the AGGRA belt transects, and most of the surgeonfishes were also small-sized. Regardless of whether or not grunts are included with the key AGRJIA carnivores, most predatory fishes were in the 1 1-20 cm size class in the two shallower reefs and somewhat larger (21-30 cm) in the two deeper reefs (Fig. 4A,B). Fish Biomass The total biomass of the AGRRA fishes (Table 7) ranged from 2,312 g/100 m^ (in D-2, which also had the fewest fishes) to 7,502 g/100 m^ (in D-1, where it was elevated by the presence of the yellowtail snapper and black durgeon). Despite their similar densities, the two shallower reefs differed somewhat in biomass due to their differing species composition, the French grunts in Sh-2 being slightly larger than the striped parrotfish in Sh-1. The key herbivores overall constituted 29% of the total fish biomass. The biomass of grunts was greater in shallow water, whereas the biomass of black durgeon, surgeonfishes and, to a lesser extent, the AGRRA-listed groupers (Epinephelus, Mycteroperca, were higher in the deeper reefs (Fig. 5). In the “all-species” surveys, the total fish biomass equaled 8,620g/100 m , of which 3,256 and 499 g/100 m respectively, were comprised of the herbivores (many of which were small-sized scarids) and carnivores (Table 6). CM E o o D) (/) (/) as E o im / y •4^ V ^ Families -A Figure 5. Mean fish biomass (grams/1 OOm^ ± sd) for AGRRA fishes in Maria la Gorda, Cuba. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. 285 Macroalgae ^ 'y The macroalgal index (data from Alcolado et al., this volume) had a significant (r = 0.52, p =0.95) inverse relationship with herbivore density, but showed no relationship with total herbivore biomass (Fig. 6). 0 50 100 150 200 M acroal^l index Figure 6. Regression plots between mean macroalgal index and (A) mean herbivore density (no. individuals/1 OOm^), (B) mean herbivore biomass (grams/ lOOm^), by site in Maria la Gorda, Cuba. DISCUSSION Relatively high values for the species diversity indices were found during the qualitative fish surveys at Maria la Gorda (Table 2). Moreover, the total number of species recorded in the “all species” transects (42 to 54, n = 6 sites) was slightly higher than the value in the Archipielago los Canarreos (mean = 42, sd = 5, n = 21 sites) recorded in 1988-1989 (Institute de Oceanologia database). The most frequently sighted and abundant of these species (blue chromis, bicolor damselfish, and bluehead wrasse) are common dominants in many other Cuban coral reefs (Claro and Garcia- Arteaga, 1994; Claro et al., 1998; Claro and Parent!, 2001). Small-sized species of parrotfishes and grunts dominated the AGRRA fishes in belt transects at the two shallower reefs. Total fish biomass in Maria la Gorda was low 286 compared to other Cuban reefs. For example, corresponding biomass values for the “all species” census were approximately 30% lower than those at comparable depths in the Archipielago los Canarreos and Archipielago Sabana-Camagiiey in 1988-1989, and less than 50% of that recorded for the Archipielago Jardines de la Reina in 1997 (Claro and Parent!, 2001). Herbivore biomass was 37% lower in Maria la Gorda than that found in 1996 in the marine reserve at the Archipielago Jardines de la Reina, where larger-sized species are more abundant (Sierra et ah, 2001). Nevertheless, the densities and biomass of herbivorous fishes in Maria la Gorda were higher than those recorded at the Archipielago los Canarreos (Table 6) and at the Archipielago Sabana-Camagiiey in 1989-1990, and even exceeded corresponding values for Martinique, Guadeloupe, and Key West, Florida (Claro and Parent!, 2001). The mean biomass of carnivorous fishes (selected groupers and snappers) was very low; values were about eight times higher in the 1988-1989 “all species” surveys at Archipielago los Canarreos (Table 6). Particularly notable was the scarcity of snappers, which are usually common in Cuban reefs (Claro and Garcia- Arteaga, 1994; Claro et ah, 1998). For example, average snapper biomass at the Archipielago los Canarreos in 1988- 1989 was thirteen times higher than that found in 1999 off Maria la Gorda. The slightly higher abundance and size of carnivores in one of its deeper reefs (D-1) may result in part from the higher relief and topographic complexity at its edge, where snappers and black durgeon tend to aggregate. Medium- to large-sized species of fishes were uncommonly scarce off Maria la Gorda. Habitat conditions (live stony coral cover, topographic relief and complexity, water quality, etc.) in 1999 seemed excellent and capable of supporting much higher fish densities. Anecdotal information and qualitative information (personal observations of the senior author) during the 1970s and 1980s indicate that Maria La Gorda’s fish populations, especially the snappers and groupers, are drastically reduced. Indeed, this area, along with Cabo San Antonio at the westernmost edge of the Biosphere Reserve, had been well known for its abundance of large fishes. Management objectives for the Biosphere Reserve are intended to protect its natural resources, to attract tourists, and to serve as reproductive replenishment reserves for downcurrent reefs; yet it is clear that fishing regulations here are poorly enforced. Large-sized fishes are vulnerable even to very low harvesting levels (Watson et al., 1997). Furthermore, fish that are subjected to spearfishing are likely to swim away from divers, which further diminishes the reserve’s touristic value. Removing large predators alters a reefs trophic balance, potentially allowing small territorial fishes and invertebrate predators of scleractinians, such as the snail, Coralliophila abbreviata, and the bristle worm, Hermodice carunculata, to proliferate. Indeed, along with blue chromis, loreto {Gramma loreto), and bluehead wrasse, the density of damselfishes at Maria la Gorda was eightfold greater than had been found in the Archipelago los Canarreos in 1988-1989, and the bicolor damselfish was 13 times more abundant (Claro, unpublished data; Table 5 for damselfishes). Individuals of the three-spot damselfish {Stegastes planifrons) were conspicuously destroying live corals in order to construct algal gardens with which to defend their territory and attract females. Fishing the larger-sized herbivorous parrotfishes can also have unintended consequences 287 when macroalgae and turf algae increase in abundance, as may have happened at Maria la Gorda (Alcolado et al., this volume). For all of these reasons, it is very important that the regulation prohibiting fishing at Maria la Gorda be strictly and effectively enforced. We also recommend increasing the no-take area protected from commercial fishing, educating the professional dive guides about the ecological and economic value of intact residential fish populations, and periodic monitoring of the local fish communities. Furthermore, the commercial and sport fishing that is carried out nearby at Cabo Corrientes during the fish spawning aggregations should be completely suspended. Establishing a carefully self-regulated “catch and release” fishery for certain non-reef fishes (tarpon, bonefish) should be considered as an alternative source of revenue for displaced fishers. ACKNOWLEDGMENTS The AGRRA organizing Committee, which facilitated financial support as described in the Forward to this volume is thanked for the invitation to collaborate in this regional initiative. We acknowledge the invaluable collaborations with, and support from. Reef Relief REFERENCES Claro, R., and L.R. Parenti 2001. The marine ichthyofauna of Cuba. Pp. 21-57. In: R. Claro, K.C. Lindeman and L.A. Parenti (eds.). Ecology of the Marine Fishes of Cuba. Washington, D.C., Smithsonian Institution Press. Claro, R., and K. C. Lindeman In press. Identification of snapper and grouper spawning aggregation sites on the insular shelf of Cuba. Proceeding of the Gulf and Caribbean Fisheries Institute 54th Annual Meeting. Claro, R., and J.P. Garcia-Artega 1994. Estructura de las comunidades de peces en los arrecifes del Grupo Insular Sabana-Camaguey, Cuba. Revista de Oceanologla y Ecologia Tropical Avicennia 2:83-107. Claro, R., J.P. Garcia-Artega, Y. Bouchon-Navarro, M. Louis, and C.Bouchon 1998. Caracteristicas de la estructura de peces en los arrecifes de las Antillas Monores y Cuba. Revista de Oceanologla y Ecologia Tropical Avicennia 8/9:69-86. Eschmeyer, W.N. (editor) 1998. Catalog of Fishes. Special Publication No. 1 of the Center for Biodiversity Research and Information, California Academy of Science, San Francisco, CA. Volumes 1-3, 2905 pp. Garcia-Arteaga, J.P., R. Claro, and S. Valle 1997. Length-weight relationships of Cuban marine fishes. NAGA, ICLARM Q. 20(l):38-43. 288 Humann, P. 1994. Reef Fish Identification, 2"^* ed. New World Publications, Inc., Jacksonville, FL, 424 pp. Watson, M., R.F.G. Ormond, and L. Holliday 1997. The role of Kenya’s marine protected areas in artisanal fisheries management. Proceedings of the Eighth International Coral Reef Symposium, Panama 11:1955-1960. Table 1. Site information for AGRRA fish surveys in Maria la Gorda, Cuba. 289 OJ B 3 O > 3 « O •u J2 "o o < E o !/3 (U =3 1/3 kH (U > (50 c ’> o c o TS -i O a c3 CC3 Cfl >> (U 3 c/3 (U X 130 l-i D X 3 C c/3 .2 ‘o (U CIh O) T — — CN 00 C3N 00 ^ !fli ON r25 cm stony corals were significantly higher in the Xcalak patch reefs than in the fore reefs, and more corals were moderately bleached at all depths off Xcalak than off Akumal. Anomalously high sea surface temperatures, having caused widespread bleaching in September 1998, possibly persisted longer at Xcalak because different exposure to hurricanes caused the southern Yucatan to cool later than the northern area. Bleaching may have rendered the Montastraea annularis species complex more susceptible to white-band disease, particularly in the Xcalak patch reefs. Partial mortality of stony corals and macroalgal abundance increased significantly in the Akumal fore reefs between 1997 and 1999, when macroalgal abundance and biomass were significantly higher than in Xcalak. Macroalgal increases off Akumal could have resulted from anthropogenic nutrient enrichment from groundwater and/or reduced grazing by Diadema antillarum and herbivorous fishes. INTRODUCTION Are coral reefs "dying" throughout the wider Caribbean? If they are, at what scale and at what rate? The Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocols are tools for rapidly quantifying spatial and temporal patterns in the condition of coral reefs. Their value is in providing species-explicit and commensurable ways of comparing reefs within regions, larger scale bioregions, or the entire, tropical western Atlantic (Ginsburg et al., 2000; Steneck et al., 2000; Kramer, this volume). The AGRRA protocols were also designed to record indicators of stress and potential agents of disturbance. We use stress to mean any loss of productivity and disturbance to mean any loss of biomass in the organism or assemblage in question (after ’ School of Marine Sciences, University of Maine, Darling Marine Center, Walpole, ME, 04573, USA. Email: Steneck@Maine.EDU ^ P.O. Box 539, Ophelia, VA 22530, USA. Email: JandL@rivnet.net 296 Grime, 1977, 1989; Steneck, 1988; Steneck and Dethier, 1994). Specifically for stony corals, transient, nonlethal bleaching indicates stress whereas disturbances are indicated by signs of disease and patterns of mortality. Early versions of the AGRRA protocols were prototyped off two villages of the Mexican Caribbean (Fig.l). Narrow fringing and fringing-barrier reefs skirt the eastern flank of the Yucatan Peninsula- where many coastal ecosystems are under assault from the rapid growth of the tourist industry (Jordan Dahlgren, 1993; Lang et al., 1998). Akumal, a resort community with a resident population of >600 near the southern end of the northerly Cancun-Tulum corridor, is surrounded by numerous large-scale developments. Xcalak, near the southern border with Belize, has a population of about 300 and is being transformed from a reliance on offshore fishing to ecotourism. Our initial assessments of these reefs were made in March and August 1997. Resurveys in March 1 999 were conducted six-seven months after the first reports of thermally induced (Fig. 2) bleaching in the Yucatan (Kramer et al., 2000), five months after Hurricane Georges passed near the northern Yucatan, and four months after Hurricane Mitch affected the southern area. We sought to determine patterns in coral mortality, whether particular species were suffering higher than average mortality, and whether mortality patterns corresponded with known stressors and disturbance agents (bleaching, hurricanes, macroalgae) or with inferred changes in the reefs’ trophic structure (i.e., reductions in herbivory). METHODS Chosen for assessment were six, moderate-relief “middle-lobe” fore reefs at depths of 1 1-18 m near Akumal (Fig. 1; Table 1). Comparable habitats were selected at 8.5-16.5 m in six fore reefs off Xcalak along with two shallow (1-3 m) patch reefs in its wider lagoon. All sites were considered to be representative of local reefs based on the knowledge and advice of residents who had dived there for years. The draft AGRRA benthos protocol was utilized by six divers in early March 1997. The following August two divers applied the revised AGRRA Version 1 procedures at two reefs off Akumal (Dick's, Las Redes). In early March 1999, when three-six divers used the AGRRA Version 2.0 protocols (see Appendix One, this volume) at all 14 reef sites, live coral cover was measured to the nearest 5 cm. The maximum diameter of stony corals (i.e., scleractinians and Millepora complanata) was approximated to the closest 5 cm with one side of the algal quadrat for scale. Bleaching was scored as “light” or “moderate,” completely bleached stony corals being absent at this time. Two species of Colpophyllia (C. breviserialis, C. natans) were recognized. Humann (1993) served as the primary field guide for other scleractinians. Stony coral recruits were not quantified. Surficial sediment was removed from the algal quadrats by gentle hand sweeping before estimating crustose coralline algal abundance. A limited number of fish surveys were made by one diver (Steneck) in 1999. The AGRRA version 2.0 fish belt-transect protocol (see Appendix One) was modified by swimming three transects, each 30 long x 2 m wide, in six of the fore reefs (Akumal- Media Luna, Dick's, Xcalak-Coral Gardens, Chimney, El Quebrado, Dos Cocos). Scarids (parrotfish) and haemulids (grunts) <5 cm in length were not counted. Although 297 Figure 2. Sea surface thermal anomalies in 1998 (for July 4, Aug. 1, Sept. 1, Oct. 6, Nov. 1 and Dec. 1) show the development of a thermal anomaly around the entire Yucatan by early September, its contraction to the south by early October and disappearance by early November. (Images from NOAA/NESDIS). 298 dives were made from 9 a.m. to 5:30 p.m., most of the fish surveys were conducted during mid-day hours (10 a.m-2 p.m.). The Fish Bite Method (see Appendix One) was used to assess herbivory in Dick's and Las Redes reefs by one diver (J. Reichman) in August 1997. Data for all species were pooled for certain sites in some analyses; in other analyses the data for certain species were pooled across sites. Variance is given either as standard deviation (sd) or as standard error (se). Changes in the AGRRA benthos protocol between March 1997 and 1999 render some comparisons (e.g., of mortality patterns and macroalgal biomass) between those two dates impossible. RESULTS Stony Corals Live stony coral cover, first measured in March 1999, averaged less than 20% for most (10/14) of the surveyed reefs (Fig. 3 A; Table 1). Variance among the fore reefs was lower off Akumal than Xcalak, where the live cover in one reef was about four times higher than at three others (37% versus ~9%, respectively). Nineteen species of “large” (>25 cm maximum diameter) stony corals were quantified in our surveys. Eighteen were found in the fore reefs where the Montastraea annularis species complex predominated, numerically comprising >50% of the surveyed colonies in 1999 (with M annularis > M. faveolata » M. franks i) and each remaining species constituting <7% (Fig. 4; Table 2). The maximum diameter of the >25 cm fore-reef corals averaged 76 cm (se=24, n=12 reefs). Colonies were somewhat smaller off Xcalak (mean=62, se=22, n=6 reefs) than in Akumal (mean=89, se=20, n=6 reefs) and there was considerable variation within some reefs, especially Akumal’s Media Luna (Fig. 3B; Table 3). A. palmata was clearly the largest of the fore-reef corals, averaging 205 cm in maximum diameter (Fig. 5; Table 2), and three (A. palmata, M. franksi, M. faveolata) of the four largest species were among the six most abundant of the fore-reef corals. Nevertheless, when abundance is expressed as percent relative coverage (mean number of colonies/m x mean maximum diameter in cm), co-dominants were M. anularis and M. faveolata, each with 10.5% of the total. Only nine species of >25 cm stony corals were recorded in the Xcalak patch reefs. M annularis was the numerical dominant with 75% of all colonies. Acropora cervicornis and Agaricia tenuifolia each had 8.5%, Porites porites comprised 6.5%, while all the remainder {A. palmata, Agaricia agaricites, Porites astreoides, Colpophyllia natans, Millepora complanata) were each <3%. Stony corals were substantially larger in the two patch reefs than in the fore reefs (Table 3), with an overall mean of 150 cm (se==14), primarily due to the higher abundance of M annularis which here averaged -165 cm in maximum diameter. 299 ^60- ^50- > S3 40- X Fore reef A. Live Stony Coral Fore reef 0 30- 20' 10- O' 1^4oa tr35a s ^3oa 25a ;-L 1 1 n I n lin n I r (U nidi^ I §2001 Q150t lOCUi 50‘ O' B. Maximum Diameter>25 cm Stony 1 I FI FT c hJ 2 'B O uS a G O Q Akumal 13 o O O 1/5 25 cm max. diameter) in March 1999 by site for Akumal (n=336) and Xcalak (n=333 in fore reefs and 196 in patch reefs). Horizontal lines = pooled mean ± standard error for each habitat. Stony Coral Condition We saw no evidence of bleaching during the 1997 surveys. In March 1999, moderate bleaching affected <3% of the >25 cm stony corals in the Akumal fore reefs and about 10% in each habitat off Xcalak (Fig. 6A; Table 3). A further 4-18% of the colonies in each reef were considered to be lightly bleached at this time. None of the >25 cm stony corals that were censused at Dick's and Las Redes reefs in August 1997 appeared diseased, although signs of white-band disease (WBD) and yellow-blotch disease were noticed in a few colonies of A. palmata and M. faveolata, respectively (Lang, personal observations). White plague was widespread in March 1999, however, with surveyed corals in 5/6 of the Akumal fore reefs, 3/6 of the Xcalak fore reefs and both patch reefs showing signs of infection (Fig. 6B; Table 3). The percentage of diseased colonies varied between areas and habitats, being ~1% in the Xcalak fore reefs, ~5% in Akumal and ~10% in the Xcalak patches. Most affected were colonies of Montastraea annularis (fore reefs-60% at Akumal and 57% at Xcalak; patch reefs-95%). 300 Relative Abundance 0 5 10 15 20 25 30 Montastraea annularis Montastraea faveolata Siderastrea siderea Montastraea cavernosa Acropora palmata Montastraea franksi Agaricia tenuifolia Diploria strigosa Agaricia agaric it es Colpophyllia natans Porites astreoides Diploria labyrithiformis Porites porities Diploria clivosa Millepora complanata Madracis mirabilis Colpophyllis breviserialis Porites furcata Figure 4. Relative abundance as % of the Akumal and Xcalak fore-reef stony corals (all >25 cm max. diameter; ^=669) in March 1999, ordered from most abundant to least abundant species. followed by M faveolata and M. franksi. Black-band disease was seen in one colony of Agaricia tenuifolia off Akumal (Table 3, footnote). Recent partial-colony mortality (hereafter recent mortality) of large stony corals averaged 1.5% (sd =2.5) in the two Akumal fore reefs that were surveyed in August 1997. Overall estimates of recent mortality in the fore-reef were higher (Akumal^. 5%, Xcalak-5.0%) in March 1999, but revealed no large-scale differences between the two areas (Fig. 7, Table 3). Old partial-colony mortality (hereafter old mortality) in the two Akumal fore reefs averaged 31.5% (se=2) in August 1997. Old mortality was somewhat elevated in fore reefs with above-average values for recent mortality, but overall estimates in 1999 were virtually identical in the two areas, averaging 27.5% (se=7.0, n=12 reefs). Recent partial mortality averaged 28% in the two Xcalak patch reefs in 1999, when mean values of old partial mortality were -40%. Hence total (recent + old) partial mortality in the patch reefs was approximately twice as high as in the fore reefs (Fig. 7; Table 3). Regardless of location, recent mortality declined steadily between 7 and 15 m in 301 Mean Diameter (cm) Montastraea annularis Montastraea faveolata Siderastrea siderea Montastraea cavernosa Acropora palmata Montastraea franksi Agaricia tenuifolia Diploria strigosa Agaricia agaricites Colpophyllia natans Porites astreoides Diploria labyrithiformis Porites porities Diploria clivosa Millepora complanata Madracis mirabilis Colpophyllis breviserialis Porites furcata Overall Mean (± SE) Figure 5. Mean maximum diameter of the Akumal and Xcalak fore-reef stony corals (all >25 cm max. diameter, n=669) in March 1999, ordered from most abundant to least abundant species (as in Figure 4). Vertical lines = mean ± standard error for all species pooled. the fore reefs (Fig. 8) and a similar, but less, dramatic depth-related decline was evident in old mortality as well. The percentage of large stony corals that were “standing dead” (entire upper surfaces dead, colony still in growth position) in March 1999 was very variable among sites (Table 3), averaging 4-5% overall in the fore reefs and about 9% in the patch reefs. However, a much higher proportion of the A. palmata were standing dead (n=16/41 colonies in the fore reefs, n=2/6 colonies in the patch reefs). Detached or fallen stony corals were rare in all the surveyed reefs. Recent mortality in Millepora and Agaricia is likely to be underreported. Their corallite structures are relatively indistinct, and soon after death the skeletons are probably scored as “old dead” by the AGRRA protocol. Bearing this caveat in mind, recent and old mortality patterns varied by species in 1999, with mean values of 3% and 22%, respectively, for the 18 fore-reef species (Fig. 9; Table 2). Values for recent mortality that were well above the average were found in M. annularis and M. faveolata, whereas lower than average values were recorded for Montastraea cavernosa, M. complanata and Colpophyllia breviserialis (Fig. 9, Table 2). Percent recent mortality of 302 o 15 10 c L. o Cu -o 14 t 12 !/! W O U e 3 ■o 'ii U c o Q u o o •o 0^ a f/i CQ Akumal o u o U V) O Q 0) "O u 3 1/3 O 3 3 U ■3 3 O ■= a o T) 3 Sh £i lU 3 JS u a Xcalak Figure 6. Percent of (A) bleaching with averages among reefs only determined for moderate bleaching and (B) white plague in stony corals (all >25 cm max. diameter) in March 1999 by site for Akumal (n=336) and Xcalak (n=333 in fore reefs and 196 in patch reefs). Mean and standard error notation as in Figure 3. colony surfaces of A. palmata was somewhat greater in Akumal (mean=6.5, se=3, n=31 colonies) than off Xcalak (mean=2, se=2, n=10 colonies). Fore-reef species with higher than average values for old (and total) partial-colony mortality were A. palmata » Agaricia tenuifolia > M. mirabilis ~ M. annularis. Mean values of recent mortality in the patch reefs were especially high for M. annularis (3 1 .5%), A. agaricites (24%), Acropora cervicornis and A. tenuifolia (each 15%), while old mortality was greatest inyl. palmata (87%), P. porites (64%), A. agaricites (55%) and M. annularis (37%). The only colony of Millepora complanata censused here was scored as having 1 00% old mortality. cd c3 C 3 2 -(U (Z1 o Q .2 2^ ‘o '4— » >> a o Vh o cd C o ■*-> o Q c/3 D *3 c/3 3 Akumal c/3 O c/3 a c/3 3 3 QJ o "O cd cd o O -3 o O T3 cd 1-1 N • ^ ti u c/3 2 3 c3 cd o CT 3 O Oh o Q S-i U Id ;-i o a u jd 5 3 00 U S Xcalak Xcalak Patch Reefs Figure 7. Total (recent and old) partial colony mortality of all stony corals (>25 cm max. diameter) in March 1999 by site for Akumal (n=336) and Xcalak (n=333 in fore reefs and 196 in patch reefs). Mean and standard error notation as in Figure 3. Depth (m) Figure 8. Plot of mean recent partial mortality of all stony corals (>25 cm max. diameter) versus water depth in March 1999 at each site off Akumal and Xcalak. Curve fit is 3rd order polynomial. 304 Montaslraea annularis Montastraea faveolala Siderastrea siderea Montastraea cavernosa Acropora palmata Montastraea franksi Agaricia tenuifolia Diploria strigosa Agaricia agaricites Colpophyllia natans Porites astreoides Diploria labyrithiformis Porites porities Diploria clivosa Millepora complanata Madracis mirabilis Colpophyllis breviserialis Porites furcata Mortality % 0 5 10 15 20 25 30 35 40 45 50 55 60 Figure 9. Total (recent and old) partial colony mortality as % of the Akumal and Xcalak fore-reef stony corals (>25 cm max. diameter, n=669) in March 1999 ordered from most abundant to least abundant species (as in Figure 4). Mean and standard error notation as in Figure 5. Algae In March 1 997, the relative abundance of macroalgae averaged ~25% in the fore reefs of both regions and ~12 % in the Xcalak patch reefs. Two years later, their relative abundance had significantly increased in the Akumal fore reefs (averaging 44%), whereas the reefs at Xcalak were essentially unchanged (Figs. 10, 11; Table 4). Macroalgal heights in 1 999 were also slightly higher in Akumal, thus macroalgal indices (an approximation of their biomass, see Table 4) were three to four times greater here than in Xcalak. Turf algae off Akumal showed a pattern opposite that of the macroalgae, having declined from 62% in 1997 to nearly 35% in 1999. However, turfs were still the predominant algal group in both habitats off Xcalak. The relative abundance of crustose coralline algae in 1999 varied from >15% in the patches to nearly 30% in the Xcalak fore reefs. Fish and Diadema antillarum The total abundance of fore-reef fishes surveyed in the belt transects in numbers/100 m was slightly higher off Xcalak (mean=34.6, se=3.1) than Akumal (mean=26.1, se=0.8). Most common on the fore reefs in 1999 were herbivores; acanthurids (mean=7.7, se=0.8) were the most abundant in Xcalak whereas scarids 305 90 ^ 80 ^70 § 60 1 50 ^40 « 30 cd ^ 10 *4-^ — (U 0^ < « _> — aj Qi 20 •■ 0 4 80 70 60 50 40 30 20 10 0 60 50 40 30 20 10 0 Algal Turf I Eore reef Fore reef Macroalgae nni Patch reef n n n n . Crustose Coralline Algae 1 1 1 n irl n II 3 03 >- 3 C 3 .2 o 3 -J 3 C o Q o o c/5 (U T3 (U C4 c/5 3 Akumal c/5 C/5 O 3 o o o c/5 o Q o c/5 3 O Z ^ HI 3 -o U l-l 3 00 Xcalak c (U ■3 3 o o U o TD 3 u X) u 3 a w >5 !U c B 15 U ii 3 42 N — 3 -d O' O C Oh 3 oa Figure 10. Relative abundance (mean ± standard deviation) of (A) macroalgae, (B) turf algae, and (C) crustose coralline algae in March 1999 by site for Akumal (n=301 quadrats) and Xcalak (n=361 quadrats in fore reefs and 136 in patch reefs). Mean and standard error notation as in Figure 3. 306 Trends in Macroalgal Abundance 1997 1999 Years Figure 11. Temporal trends (mean ± standard error) in relative abundance of macroalgae in the Akumal and Xcalak fore reefs and the Xcalak patch reefs for March 1997 and March 1999. The increase in abundance in the Akumal fore reefs was the only significant difference (ANOVA, p <.0001) between the surveys. 20 I 15 0) a I 10 u. £ 5 Figure 12. Density (mean no. fish/lOOm^ ± standard error) of AGRRA fishes in the Akumal and Xcalak fore reefs in March 1999. Other = Bodianus rufus, Caranx ruber, Microspathodon chrysurus. 307 (mean=8.3, se=0.6) were predominant off Akumal (Fig. 12). Parrotflsh (scarids) were slightly larger in the southern fore reefs: average lengths of the five species present in belt transects were 16.33 cm (se=1.15) off Xcalak versus 14.64 cm (se=0.99) off Akumal. Large-sized groupers (serranids) were absent, but smaller species such as coneys (Epinephelus fulvus), red hinds {E. guttatus) and graysbys {E. cruentatus) collectively averaged between 1.4 and 3.3/100 m^ off Akumal and Xcalak, respectively. Bite rates for grazing fishes (acanthurids, scarids, Microspathodon chrysurus) averaged 175/m^/hour (se=160) at Dick’s and Las Redes reefs in Akumal in August 1997; similar results were obtained at two other fore-reef sites in this area. The long-spined sea urchin, Diadema antillarum, was rarely seen in fore-reef habitats and none were counted in any of the transects in March or August 1997. However, a few (3. 5/ 100m ) were found in one of the Xcalak patch reefs in March 1999 (Table 4). DISCUSSION Are coral reefs "dying” throughout the Caribbean? To address this question, we need baseline information against which comparisons can be made. Our data for reefs in two areas of the Yucatan can contribute to a baseline, but no single study can address this question because each is too limited in its spatial and temporal coverage. Comparisons require combining our results with other AGRRA studies throughout the Caribbean as discussed by Kramer (this volume). It must also be noted that rapid assessments are little more than “snapshots” of reef condition. We have no way of determining variability in time outside of our survey periods. Although assessments were conducted in the same reefs and during the same weeks of March 1997 and 1999, we cannot say if a month earlier or later the patterns would have remained the same. However, populations of key fish groups at least are unlikely to exhibit short-term increases, particularly given ongoing spearfishing around Akumal (personal observations) and the as-yet unregulated poaching that occurs in Xcalak (S. Redman, personal communication). Nevertheless, many of the AGRRA indices for 1999 are suggestive that the Yucatan reefs we studied may be in decline. For example, average live stony coral cover overall was <20% (Fig. 3 A; Table 1) and, for the Akumal-area fore reefs, was somewhat lower than Munoz-Chagin and de la Cruz-Aguera’s (1993) previous area-based estimate (18% versus 23%, respectively). Population densities of the AGRRA fishes were low in the seven examined fore reefs (Fig. 12), and the abundance of macroalgae had increased significantly over a period of two years off Akumal (Fig. 11). Mean values of recent and old partial mortality for surveyed corals were extremely high in the two Xcalak patch reefs (Fig. 7). Partial (recent and old) mortality values for the >25 cm stony corals were similar in the fore reefs off Akumal and Xcalak (Fig. 7) in March 1999. Nonetheless, the underlying causes for their mortality may differ since numerous potential stressors and disturbance agents, such as hurricanes, fishing, groundwater input and outbreaks of disease, complicate simple explanations. Below, we discuss how the sudden “pulse” 308 events of 1998 may have interacted with “chronic” stresses and disturbances in the region from 1997 to 1999 to produce the patterns of mortality we recorded. Recent Pulse Events: Mass Bleaching and Hurricanes of 1998 High temperatures can induce stony corals to bleach, and they will die if the thermal condition persists (e.g., Fitt et al., 2001). Over the past century tropical oceans have been warming which, together with El Nino-related periods of elevated temperatures, have contributed to an increased frequency of coral bleaching (Glynn, 1993; Brown, 1997). The strong El Nino of 1997-1998 resulted in unusually warm sea surface temperatures and widespread lethal bleaching of stony corals worldwide (Wilkinson, 2000). By September 1, an enormous area that was nearly +1.5° C above the average monthly high temperature had developed throughout the western Caribbean (Fig. 2; see also Kramer et al., 2000; Kramer and Kramer, in press). As September is the warmest month in the Caribbean, a warm anomaly at this time is most likely to exceed the thermal thresholds (sensu Fitt et al., 2001) of many stony corals. Indeed local divers observed widespread bleaching at Akumal (D. Brewer and S. Slingsby, personal communication) and Xcalak (T. Biller, personal communication). Subsequent spatial and temporal patterns of sea-surface cooling are consistent with hurricane activity in the area. Hurricanes propagate large-amplitude, low-frequency storm swells that can mix warm surface water layers with deep cooler water. Two hurricanes traversed the western Caribbean in 1998. On September 25, Georges passed over the northwest coast of Cuba, adjacent to the northeast Yucatan coast, as a Category 2 hurricane with sustained winds of 169 km/hour and continued northwest toward the Mississippi delta. The satellite image for October 6 shows an entirely different thermal anomaly pattern from that of the previous month, with most of the western Caribbean north of 19°N latitude, including Akumal, being only about +0.5° C above the monthly average (Fig. 2). This rapid cooling resulted in only transient rather than lethal bleaching over much of the Mexican Caribbean. Nevertheless, a warm region of nearly +1.5° C persisted throughout the southwestern Caribbean from southern Mexico (including Xcalak) through Belize and Honduras to Colombia. Then Hurricane Mitch developed off the Colombian coast on October 24. Over the next four days it slowly built to a Category 5 hurricane with sustained winds of 290 km/hour as it stalled in the Gulf of Honduras adjacent to Belize. Mitch generated 1 0 m-high storm swells that crested and broke on the Xcalak reefs (T. Biller, personal communication). Three days later, the November 1 satellite image recorded no thermal anomalies anywhere in the western Caribbean. In fact, the only remaining warm spots were localized off northwestern Cuba, in the southweastem Gulf of Mexico and areas in the ordinarily hurricane-free region from south of Puerto Rico to Venezuela. That moderately-bleached >25 cm stony corals were more common in March 1999 off Xcalak (both habitats) than in the Akumal fore reefs (Fig. 6) is consistent with the pattern of a longer-duration thermal event in the south having created greater stress and allowed less time for recovery. Reef organisms would have experienced higher temperatures (and presumably higher solar irradiance) in the shallow (1-2 m), calm waters of the Xcalak lagoon than in its offshore fore reefs. Lethal bleaching probably 309 contributed to the extremely high values of recent partial mortality (~27% of colony surfaces) found in the patch reefs, and thereby reduced the percentage of colonies still showing residual levels of light or moderate bleaching in March 1 999 to approximate those in the fore reefs nearby (Table 3). Bleaching-induced mortality of A. tenuifolia was first reported in the Yucatan in October 1998 (Kramer et al., 2000). As the transition from what appears to be “recent” to what is scored as “old” probably occurs relatively quickly in species of Agaricia, we also suspect that at least some of the comparatively high values for old mortality in the >25 cm fore-reef colonies of this species (Table 2; Fig. 9) had occurred as a result of the 1998 bleaching event. Mexican reefs sustained relatively little physical damage from hurricanes in 1998, as evidenced by the scarcity of detached and fallen colonies in the March 1 999 transects (see also Kramer et al., 2000). That macroalgal blooms were conspicuous off Akumal immediately following Hurricane Mitch (S. Slingsby, personal communication) is consistent with the significant increase in their abundance recorded in the Akumal fore reefs (but not at Xcalak) between 1997 and 1999 (Fig. 11). Chronic Events: Trends and Agents of Mortality Recent and old partial-colony mortality of the >25 cm stony corals increased significantly between 1997 and 1999 in the two Akumal fore reefs for which commensurable protocols were applied (Fig. 13; T-test, p <0.05). We do not know how long “recently dead” coral skeletons generally remain in that category before being recorded as “old dead” according to criteria in the AGRRA protocol (Appendix One). However, colonies of the abundant Montastraea annularis species complex and of Acropora palmata frequently retain surficial corallite structures for months after death, particularly when grazing pressures are low (personal observations). Thus we suspect that a majority of the increase in recent mortality recorded in March 1999 was the result of events occurring since the previous summer whereas, apart from A. tenuifolia, the higher values of old mortality reflected cumulative impacts of disturbances which had occurred since the 1997 surveys. Recent Mortalitv Old Mortalitv 1997 1999 1997 1999 Figure 13. Temporal trends (mean ± standard error) in recent and old partial colony mortality of stony corals (all >25 cm max. diameter; n=52 in August 1997 and 93 in March 1999) in Las Redes and Dick's reefs off Akumal. Comparable data do not exist for March 1997 because of changes in the AGRRA protocol. 310 Outbreaks of disease (and potential disease states, sensu Richardson, 1998) appear to have increased in diversity and severity throughout the Caribbean in recent decades, and had probably contributed to the above-average levels of partial mortality observed in some species in March 1999 (Fig. 9; Table 2). In particular, the stress of the 1998 bleaching event might have increased the susceptibility of Montastraea annularis and related species to white plague. A high percentage of diseased colonies were also reported throughout the Mesoamerican reef system after the 1998 bleaching event, although the incidences (at least in Belize) had decreased by 2000 (Kramer et al., 2000; Kramer and Kramer, in press). WBD is apparently specific to species of Acropora and it spreads relatively slowly (Gladfelter, 1982). A few live A. palmata can still be found in fore-reef habitats to depths of about 1 5 m and we have seen signs of WPD in some of the colonies off Akumal. Thus disease might have caused the above-average levels of both old and recent partial mortality observed in this species in 1999 (Table 2; Fig. 9). High macroalgal biomass often correlates with mortality of stony corals (Lewis, 1986; Hughes, 1994; Steneck, 1994), and may result in reduced growth (Lewis, 1986; Lirman, 2001) as well as decreased recruitment of juvenile scleractinians (Birkeland, 1977). By having provided substrata for algal colonization and growth, the skeletons of A. palmata (Aronson and Precht, 2000) and other corals killed by disease or lethal bleaching (such as the M. annularis species complex and A. tenuifolia, respectively) may have facilitated the expansion of macroalgae in the Yucatan’s fore reefs. In turn, overgrowth of live coral polyps by macroalgae may have contributed to the relatively high values of recent mortality found in the Akumal population of A. palmata. Proliferations of macroalgae can also occur as a result of increases in nutrients (e.g., Lapointe, 1997) and/or declines in herbivory (e.g., Hughes, 1994; Steneck 1994). Hasty development of coastal infrastructures, locally high human populations, inadequate sewage treatment (especially in the pueblos), and the hydrography of the Caribbean coast of Mexico make it particularly vulnerable to nutrient enrichment. Groundwater discharged from a ramified underground river complex of cenotes permeates much of the northern Yucatan coast. At Akumal, where discharges are commonly observed in the deep, shelf-edge reefs (D. Brewer, personal communication), nutrients in groundwater potentially pose an elevated risk because of the close proximity of the reefs to the shore. Although herbivorous acanthurids and scarids were among the more common fishes quantified in the fore-reefs off Akumal and Xcalak (Fig. 12), their population densities were low (combined means of 16/lOOm^) compared to many other areas in the wider Caribbean (Kramer, this volume). The 1997 fish herbivory rates off Akumal resembled those found at 10 m depths in 1988 on the severely overfished reefs of Jamaica, but were less than 10% the rates which had been obtained that year in St. Croix (Steneck, 1994). The key herbivore, Diadema antillarum, was very rare, particularly in the fore reefs (Table 4), and other grazing echinoids, such as Echinometra viridis and Tripneustes ventricosus (e.g., Aronson and Precht, 2001; Edmunds and Carpenter, 2001), were notably scarce (personal observations). Hence, low levels of herbivory may well have contributed to the high relative abundance and biomass of macroalgae, particularly off Akumal. 311 As populations of Diadema have begun to recover in some Caribbean reefs, macroalgae are becoming less abundant (personal observations). Edmunds and Carpenter (2001) have recently reported an upsurge in coral recruitment at shallow (<9m) fore reefs near Discovery Bay, Jamaica in which densities of Diadema had increased. Since 1999, populations of Diadema have similarly rebounded in lagoonal patch reefs off both Akumal (R. Roy, personal communication) and Xcalak (S. Redman, personal communication). Should similar increases occur in the Akumal fore reefs, the build-up seen in its macroaglal populations might be reversed at least partially. ACKNOWLEDGMENTS We received financial, logistical and technical support from many sources. Most of the benthic surveys were conducted by students of the University of Maine's School for Marine Sciences “Coral Reefs” course, and we particularly thank E. Annis, A. Calhoon, K. Eichorst, M. Grant, M. Hunter, D. McNaught, A. Palma, J. Vavrinec, C. Wilson and S. Zimsen. Additional help with the data collection was provided by T. Biller, P. Kramer, J. Reichman, M. Ruiz, and S. Slingsby. Invaluable information was shared by G. Arcila, S. Redman, R. Roy, M. Ruiz, C. Shaw and S. Slingsby. Logistical support of boats, scuba fills and other help came from D. Brewer and G. Arcila of the Akumal Dive Shop as well as from Tom Biller of Aventuras Chinchorro (now XTC Dive Center). Data entry and analysis was assisted primarily by J. Stevens with help from R. Russell, T. Kim, M. Cusack and P. Kramer. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. Additional funding was provided by the Texas Memorial Museum. To all we are grateful. REFERENCES Aronson, R.B. and W.F. Precht 2001 . White-band disease and the changing face of Caribbean coral reefs. Hydrobiologia 460:25-38. Birkeland, C. 1977. The importance of rates of biomass accumulation in early successional stages of benthic communities. Proceedings of the Third International Coral Reef Symposium, Miami 1:15-21. Brown, B.E. 1997. Coral bleaching: causes and consequences. Coral Reefs 16, Suppl.:S129-S138. Cho, L. and J. Woodley 2000. Diadema antillarum: a facilitator of recovery on the refs of Discovery Bay, Jamaica. Proceedings of the Ninth International. Coral Reef Symposium Abstract, p. 79. Edmunds, P.J., and R.C. Carpenter 2001 . Recovery of Diadema antillarum reduces macroalgal cover and increases abundance of juvenile corals on a Caribbean reef. Proceedings of the National Academy of Sciences 98:5067-5071. 312 Fitt, WK, B.E. Brown, M.E. Warner, and R.P Dunne 2001. Coral bleaching: interpretation of thermal tolerance limits and thermal thresholds in tropical corals. Coral Reefs 20:51-65. Ginsburg, R., P. Alcolado,. E. Arias, A. Bruckner, R. Claro, A. Curran, A. Deschamps, A. Fonseca, J. Feingold, C. Garcia-Saez, D. Gilliam, S.D. Gittings, A. Glasspool, G. Horta-Puga, K. Klomp, P.A. Kramer, P.R. Kramer, Z. Leao, J. Lang,, C. Manfrino, R. Nemeth, C. Pattengill-Semmens, P. Peckol, J. Posada, B. Riegl, J. Robinson, P. Sale, R. Steneck, J. Vargas, and E. Villamizar 2000. Status of Caribbean Reefs: Initial results from the Atlantic and Gulf Rapid Reef Assessment (AGRRA) Program. Proceedings of the Ninth International. Coral Reef Symposium Abstract, p. 2 1 1 . Gladfelter, W.B. 1982. White-band disease in Acropora palmata: implications for the structure and growth of shallow reefs. Bulletin of Marine Science 32:639-643. Glynn, P.W. 1993. Coral reef bleaching: ecological perspectives. Coral Reefs 12.- 1-1 7. Grime, J.P. 1977. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:1169-1194. Grime, J .P. 1989. The stress debate: symptom of impending synthesis? Biological Journal of the Linnean Society 37:3-17. Hughes, T.P. 1994. Catastrophes, phase shifts, and large-scale degradation of a Caribbean coral XQQt Science 265:1547-1551. Humann, P. 1993. Reef Coral Identification. New World Publications, Inc., Jacksonville, FL, 252 pp. Jordan Dahlgren, E. 1993. Atlas de Los Arrecifes Coralinos del Caribe Mexicano. I. El Sistema Continental. Centro de Investigaciones de Quintana Roo Institute de Ciencias del Mar y Limnologia Mexico, D. F. 110 pp. Kramer, P.A., and P. Richards Kramer In press. Transient and lethal effects of the 1998 coral bleaching event on the Mesoamerican Reef System. Proceedings of the Ninth International Coral Reef Symposium. Kramer, P.A., P.R. Kramer, A. Arias-Gonzalez, and M. McField 2000. Status of coral reefs of northern Central America: Mexico, Belize, Guatamala, Honduras, Nicaragua and El Salvador. Pp: 287-313. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science. Cape Ferguson, Queensland and Dampier, Western Australia. Lang, J., P. Alcolado, J.P. Carricart-Ganivet, M. Chiappone, A. Curran, P. Dustan, G. Gaudian, F. Geraldes, S. Gittings, R. Smith, W. Tunnell, and J. Wiener 1998. Status of coral reefs in the northern areas of the Wider Caribbean, Pp. 123-134. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 1998, Australian Institute of Marine Science. 313 Lapointe, B.E. 1997. Nutrient thresholds for eutrophication and macroalgal overgrowth of coral reefs in Jamaica and southeast Florida. Limnology and Oceanography 42:1119-1131. Lewis, S.M. 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Lirman, D. 2001. Competition between macroalgae and corals: effects of herbivore exclusion and increased algal biomass on coral survivorship and growth. Coral Reefs 19:392-399. Munoz-Chagin, R.F., and G. de la Cruz-Aguera 1993. Corales del arrecife de Akumal, Quintana Roo. Pp. 761-771. In: S.I. Salazar- Vallejo and N.E. Gonzalez (eds.), Biodiversidad Marina y Costera de Mexico. Com. Na. Biodiversidad CONABIO y CIQRO, Mexico, D.F. Richardson, L.L. 1998. Coral diseases: what is really known? Trends in Ecology and Evolution 13:438-442. Smith, J.E., C.M. Smith, C.L. Hunter 2001. An experimental analysis of the effects of herbivory and nutrient enrichment on benthic community dynamics on a Hawaiian reef Coral Reefs 19:332-342. Steneck, R.S. 1988. Herbivory on coral reefs: a synthesis. Proceedings of the Sixth International Coral Reef Symposium, Australia 1 :7-49. Steneck, R.S. 1994. Is herbivore loss more damaging to reefs than hurricanes? Case studies from two Caribbean reef systems (1978-1988). Pp. 220-226. In: Ginsburg, R.N. (compiler). Proceedings of the Colloquium on Global Aspects of Coral Reefs: Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Steneck, R.S. and M.N. Dethier 1994. A functional group approach to the structure of algal-dominated communities. Oikos 69:476-498. Steneck, R.S, R.N. Ginsburg, P. Kramer, J. Lang, and P. Sale 2000. Atlantic and Gulf Rapid Reef Assessment (AGRRA): a species and spatially explicit reef assessment protocol. Proceedings of the Ninth International. Coral Reef Symposium Abstract, p. 220 Wilkinson, C. 2000. Executive summary. Pp. 7-17. In: C. Wilkinson (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science, Cape Ferguson, Queensland and Dampier, Western Australia Table 1. Site and transect information for AGRRA surveys in March 1999 off Akumal and Xcalak, Mexico. 314 Table 2. Abundance, size and mortality of large (>25 cm diameter) reef-building corals in March 1999, by species in fore reefs off Akumal and Xcalak. 315 Table 3. Size and condition (mean ± standard deviation) of all stony corals (>25 cm diameter) in March 1999, by site off Akumal and Xcalak. 316 CQ VO ^ ^ ^ o ^ CN O O > c m m (N (N (N CN I (N CN 2 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 +1 1 +1 +1 o O NO in O in in m in o in in in 1 in in o ON NO rn o ON o 1 ON V (N (N m (N > -o O 0) 0 0^ 1 ° •§. iz 2! C o •S 25 cm diameter) stony corals at ~10 m on fore reefs in the central and southern areas, respectively, of Quintana Roo, Mexico. Reductions in live stony coral cover (from -25% to -12% in <10 years) and high recent partial-colony mortality (7-27.5%) are indications of declining reef conditions. Diseases in the five more northerly reefs, as well as tissue loss from the 1998 El Nino Southern Oscillation (ENSO) bleaching event and/or from on-going bleaching during June-July 1 999 in the three southernmost reefs, appeared responsible for much of the recent mortality. Although turf algae predominated everywhere, macroalgae were relatively more abundant in the five more northerly reefs (four of which are in a reserve where herbivorous fishes currently are less numerous than further south). Additional perturbations associated with tourism development in the southern area could result in a loss of resilience of these coastal reefs. INTRODUCTION The greatest reef development in Mexico is found in its Caribbean reefs which are principally of the fringing type. Extending for over 350 km along the coast of the state of Quintana Roo (Gutierrez et ah, 1995), they comprise the northernmost part of the Mesoamerican Reef System which continues through the neighboring countries of Belize, Guatemala and Honduras. Tourism has become the principal economic activity along the Quintana Roo coast representing -75% of the state’s 1999 Gross Internal Product (EMEGI, 2000). At ' Laboratorio de Ecologla de Ecosistemas de Arrecifes Coralinos, Centro de Investigacion y de Estudios Avanzados del IPN, Carretera Antigua a Progreso Km 6, AP 73, Merida, Yucatan, Mexico. Emails: mruiz@mda.cinvestav.mx, earias@mda.cinvestav.mx Ecole Pratique des Hautes Etudes, Laboratoire d’Ichtyoecologie Tropicale et Mediterraneenne, ESA 8046 CNRS, Universite de Perpignan, 66860 Perpignan cedex France. 320 present this activity is concentrated throughout the northern coastal area, such as at Isla Mujeres, Cozumel, and along the Cancun-Tulum Corridor. However, tourism infrastructure is either planned or developing in some southern areas of the state, particularly between Pulticub and Xcalak (Daltabuit Godas, 1999). Specific threats associated with this development include ecologically inappropriate forms of coastal development, diving, pollution from solid and liquid wastes, vessel traffic, and fishing (Daltabuit Godas, 1999). Another important aspect of the Mexican Caribbean is that a number of protected areas have been created with a view towards direct or indirect protection of its coral-reef ecosystems. The largest among these is the Sian Ka’an Biosphere Reserve established in 1986 between 19°05' N and 20°06' N in the central area of the coast. This reserve is protected by regulations restricting commercial fishing activities, aquatic sports, and development. Hence coral reefs are severely threatened by intense tourism activity in the northern area, protected in the central area, and potentially at risk in the south. Consequently, reef “health” within each area of the Mexican Caribbean may vary as a function of these differing levels of anthropogenic input. The aim of this study was to evaluate benthic reef condition in the spur-and-groove habitat of the fore-reef slope zone in the central and southern areas using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) methodology. METHODS The study area extends along approximately 230 km of the Quintana Roo coast (Fig. 1) and includes eight fringing reefs, four of which are located in the central area within the Sian Ka’an Biosphere Reserve. On the basis of morphology and degree of development, the Quintana Roo reef system has been divided into 1 6 subregions (ICRI, 1998). Three reefs (Boca Paila, Punta Yuyum, Punta Allen) are in the Sian Ka’an North subregion which is characterized as having a narrow continental shelf with steep slopes, “intermediate amounts” of freshwater input, narrow to “moderately wide” lagoons and, generally, highly developed fore-reef spurs and grooves. Coral cover averaged nearly 25% in the early 1990’s when 42 species of scleractinian eorals were found in this subregion (Gutierrez et al., 1993; ICRI 1998). These three reefs are distinguished from the remainder in having submerged crests and lagoons that are open to the ocean, and by their close proximity to coastal bays (Ascension and Espiritu Santo) and lagoons (Campechen, Boca Paila and San Miguel). Reef structures are more highly developed and complex overall on the five southernmost reefs (one of which, Tampalam, is located within the Biosphere Reserve in the central geographic area). The extreme northern portion of their reef crests runs into the shore, thereby enclosing and protecting large seagrass meadows landward of the reef fiats. Tampalam and El Placer in the Sian Ka’an South subregion have narrow reef crests and slightly lower values for eoral cover and scleractinian species number (22% and 41, respectively). Two reefs (Mahahual and Xahuayxol) in the Mahahual subregion, which are in the “shadow” of an offshore bank (Banco Chinchorro), have particularly well- developed fore-reef terraces with high-relief spurs and grooves, and back reefs that 321 extend almost to the beach. The highest values for coral cover (nearly 28%) and species richness (46) occurred in this subregion. Xcalak, in the Xcalak Trench subregion, has a well-developed lagoon and breaker zone and well-developed spurs and grooves on the fore reef and is somewhat under the shadow of Banco Chinchorro. Coral cover here was nearly 25% but there were only 33 species of corals in the early 1990s. A characteristic of all the studied reefs is that the fore reef develops as a series of “strips” (spurs and grooves) that run parallel to the coast and are usually separated from each other by intervening sand channels. Hence it is possible to differentiate between an inner fore reef (at depths ranging from 6 m to as much as 25 m), an outer fore reef (from ~1 5 m to as deep as 40 m), and a deep fore reef (from ~35 to ~50 m) terminating in a shelf-edge reef at the continental margin (ICRI, 1998). It has been suggested that the Mexican Caribbean reef system has been molded by the high frequency of tropical storms and hurricanes crossing this area, particularly near the Belizean border in the southern geographic area and between Cozumel and the Yucatan Channel in the northern area (ICRI, 1998). The last significant storm that passed near our study areas was Hurricane Mitch in October 1998; nevertheless, there are just three descriptions of its effects at Mahahual (Garza-Perez, 1999; Garza-Perez et al., 2000; Bastida-Zavala, 2000). Similarly, there is only one formal report of the 1998 mass bleaching event, which is an important recent disturbance. Its effects were documented at Xcalak, Mahahual, Tampalam, and further north at Akumal by Garza-Perez (1999), who continued to assess the condition of the Mahahual reef through 2000 (Garza-Perez, in preparation). This study was conducted in the fore-reef spur-and-groove habitat at an average depth of 10 m where stony coral cover in central-southern Quintana Roo is highest (Gutierrez et al., 1993; Gutierrez et al., 1995). We did not survey the outer-reef crest, another important habitat, for logistical reasons and because it is not well-developed on all the study reefs. We established a hierarchical survey design based on three spatial scales: geographic area (central and southern), reefs (four in each area), and subreefs (three per reef). Reefs that were known to exhibit maximum spur-and-groove development were chosen on the basis of previous field work and from published descriptions (Gutierrez et al., 1993; Jordan, 1993) and were thus selected on the basis of both strategic and representative criteria (see Appendix One). The distance between reefs, which were named after the nearest town or coastal feature (Fig. 1), was approximately 20-30 km, except for Tampalam which is ~75 km south of Punta Allen. To avoid pseudoreplication (i.e., Oxley 1997), the north (N), center (C) and south (S) subreef for each reef were positioned 0.9-1 km apart. This separation distance was chosen to prevent large geomorphological changes between the subreefs of any given reef. Their geographical coordinates were established by GPS (Table 1). Borders delimiting each subreef were temporarily established within which surveys were performed on haphazardly positioned transects over the spur structures in the inner spur-and-groove zones. Fieldwork was carried out by two observers (Ruiz-Zarate on the transects and Hemandez-Landa on the quadrats) between June and October 1999 (Table 1), using the AGRRA Version 2.2 benthos protocols (see Appendix One, this volume). The sizes of the individually surveyed stony corals (scleractinians and Millepora spp.) were measured to the nearest cm. Crustose coralline algae were exposed by vigorous removal of 322 sediment and pushing apart any macroalgae and species of scleractinians that are small as adults were omitted from the counts of tiny (<2 cm diameter) corals. Training exercises were conducted in the field prior to performing the surveys. Humman’s (1993a, 1998b) field guides aided in the identification of organisms. Averages were computed for each spatial scale (subreef, reef, and area). Correlation and regression analyses were then applied to find statistical tendencies. RESULTS Stony Corals Four morphological forms of Agaricia agaricites (A. agaricites agaricites, A. agaricites carinata, A. agaricites danai, A. agaricites purpurea) that were recognized in this study have not previously been discriminated in the Mexican Caribbean and, for matters of comparison, they were considered as a single species {Agaricia agaricites). A total of 24 species of “large” stony corals (colonies >25 cm maximum diameter) were recorded in our transects, representing nearly 60% of the 47 species listed in central- southern Quintana Roo by Beltran-Torres and Carricart-Ganivet (1999). Of these 47 species, 37 can be expected to occur in depths of about 10 m (Cuadro 2; Beltran-Torres and Carricart-Ganivet, 1999; personal observations). Twenty-six are reported to grow larger than 25 cm in diameter (after Humann, 1993), seven to be smaller than 25 cm, and three of the four species for which no sizes were given {Porites furcata, Montastraea faveolata, M. franks i) are well known to exceed 25 cm in diameter while the fourth, Porites divaricata, is usually small. Included in our surveys were 23 (79 %) of the 29 large species known to occur in this habitat and one species {Madracis decactis) that Humann (1993) had characterized as being less than 25 cm in diameter. Four species numerically constituted 50% of the total: Montastraea annularis > M. faveolata > M. cavernosa > Diploria strigosa (Fig. 2). Species with relative frequencies of 70-100% (i.e., recorded at 17-24 of the subreefs), in order of decreasing frequency, were D. strigosa, M. faveolata, M. cavernosa, M. annularis, Siderastrea siderea, Colpophyllia natans, D. labyrinthiformis, Porites astreoides and M. franksi. The two reefs (Xcalak, El Placer) with the lowest mean live stony coral cover (~9%), and Mahahual, with the highest mean cover (-17%), were all in the southern area (Table 1). Mean live stony coral cover for the four reefs in the central area was virtually identical to that of the four southern reefs and averaged 12.0% overall (sd = 6.2, n = 3 1 1 transects). The mean density of large (>25 cm diameter) stony corals varied from 1.5- 6.0/10 m transect. 323 Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>25 cm diameter) at 8-12 m on fore reefs in central-southern Quintana Roo, Mexico. Other = Agaricia tenuifolia, Meandrina meandrites, Porites furcata, P. porites, Acropora cervicornis, Madracis mirabilis, M. decactis, Millepora complanata, M. alcicornis, MycetophylUa ferox, M. danaana, Dendrogyra cylindrus, Eusmilia fastigiata, and standing dead colonies of Diploria, Montastraea and Agaricia. The diameters of large stony corals were smallest in El Placer 48 cm) and largest in Boca Paila (~ 65 cm) (Table 2). Stony corals were slightly larger overall in the central area than in the southern area; the mean diameter for both areas combined was 56.3 cm (sd = 40.6, n = 1,218 corals). The size-frequency distributions, both as maximum diameter and as maximum height, of the five most abundant large species in each geographic area are summarized in Figures 3A, B. Colonies of M. faveolata, which were more abundant in reefs of the central area, constituted the largest corals in both areas. Apart from M faveolata, most of the surveyed colonies were in small to intermediate size classes (30-50 cm for maximum diameter, 10-60 cm for maximum height). The mean density of stony coral recruits in the different reefs varied from 0.1- 0.4/0.0625 m , averaging 0.2 /0.0625 m in both geographic areas (Table 3). In many cases, it was not possible to identify the coral recruits to the species level. The three most important genera, in order of relative abundance, were Agaricia, Siderastrea and Porites (Fig. 4). Only one individual of the herbivorous echinoid Diadema antillarum was found in a belt transect in one subreef (Tampalam N). 324 A Central Qunitana Roo Montastraea annularis n = 63 10 30 50 70 90 110 130 150 170 190 210 230 250 Montastraea faveolata n = 106 40 : 20 1 0 4_n ^_E1. j] , j] , J1 , ■ ,■ ^ I — I- 10 30 50 70 90 110 130 150 170 190 210 230 250 Montastraea cavernosa n = 95 I" 10 30 50 70 90 110 130 150 170 190 210 230 250 Diploria strigosa n = 55 10 30 50 70 90 110 130 150 170 190 210 230 250 Agaricia agaricites species complex n = 82 40 20 3Ll iLi 10 30 50 70 90 110 130 150 170 190 210 230 250 Size Class (cm) Figure 3A. Size-frequency distribution as diameter (dark bars) and height (white bars) of colonies (>25 cm diameter) at 8-12 m on fore reefs in central Quintana Roo, Mexico. 325 B Southern Quintana Roo Montastraea annularis n = 119 40 I 20 1 0 i -iti ■Il.tf , jg, n“'“T r ^ I “ I I r 10 30 50 70 90 110 130 150 170 190 210 230 250 Montastraea faveolata n = 65 40 J 20 J 10 30 50 70 90 110 130 150 170 190 210 230 250 Montastraea cavernosa n = 51 I" 10 30 50 70 90 110 130 150 170 190 210 230 250 b Diploria strigosa n = 69 40 20 0 10 30 50 70 90 110 130 150 170 190 210 230 250 Siderastrea siderea n = 54 10 30 50 70 90 110 130 150 170 190 210 230 250 Size Class (cm) Figure 3B. Size-frequency distribution as diameter (dark bars) and height (white bars) of colonies (>25 cm diameter) at 8-12 m on fore reefs in southern Quintana Roo, Mexico. 326 Figure 4. Species composition and mean relative abundance of all stony coral recruits (>2cm diameter, excluding species that are small as adults) at 8-12 m on fore reefs in the central and southern Mexican Caribbean. Other = Stephanocoenia intersepta, Montastraea spp., Madracis spp., Meandrina meandrites, Manicina areolata, Acropora cervicornis, Isophyllia sinuosa, Colpophyllia natans, Millepora alcicornis and a tiny unknown scleractinian. Figure 5. Frequency distribution as % of old partial colony mortality (black bars) and recent partial colony mortality (white bars) of all stony corals (>25 cm diameter) at 8-12 m on fore reefs in central- southern Quintana Roo, Mexico. 'ill Coral Condition Recent partial-colony mortality (hereafter recent mortality) was relatively high, exceeding 15% in 70 % (17/24) of the subreefs, with only one reef each in the central (Tampalam) and southern (Xcalak) areas having mean values of <15%. Levels of recent mortality in the two areas were very similar, and the overall mean was 18.0% (sd = 20.0, n = 1146 stony corals). The two reefs (Mahahual, Xahuayxol) with the lowest mean levels of old partial- colony mortality (hereafter old mortality), as well as the reef with the highest mean old mortality (Xcalak) were all in the southern area. Mean values of old mortality overlapped at the area scale and the overall mean of 19.3% (sd. = 25.4, n = 1218 stony corals) was close to that for recent mortality. Nonetheless, the most frequent percentage class category for both recent and old coral mortality was 0-10% (Fig. 5). Total (recent + old) partial-colony mortality in the reefs ranged between about 30% at Xahuayxol in the southern area to 43% at Punta Yuyum in the central area. The percentage of “standing dead” stony corals (i.e., 100% mortality of upward facing surfaces of colonies still in the original position of growth) exhibited two pronounced peaks: between Punta Yuyum and Punta Allen in the central area (6-16% of all colonies, n=6 subreefs), and at Xcalak in the southern area (12-24% of colonies, n = 3 subreefs). Acropora palmata accounted for about half to two-thirds of the standing dead corals in Punta Allen (10/18) and Xcalak (13/19) but Diploria represented half (10/19) of the standing dead colonies in Punta Yuyum. Corresponding values for standing dead on the remaining subreefs ranged from 0-6% (Table 2). Except for two subreefs in the central area (Punta Allen N, Tampalam N), bleaching was restricted to the southern geographic area, principally between Mahahual N and Xahuayxol C (Fig. 6). “Partly bleached” was the most common condition in these five subreefs affecting 18-35% of the large stony corals. Only in Mahahual, where the highest percentages of stony corals with bleached tissues were found, were all three bleaching categories (pale, partly bleached, bleached) recorded (Fig. 6). Stony corals: bleaching condition n=l 19 Subreef Figure 6. Percentages of all stony corals (>25 cm diameter) affected by bleaching (as pale, partly bleached, bleached) at 8-12 m on fore reefs in central-southern Quintana Roo, Mexico. 328 Diseased stony corals were relatively rare in the three most southerly of the reefs (being 0% at Xahuayxol) (Table 2). Percentages of diseased corals were highest in Boca Paila (-30%) and Tampalam (25%) in the central geographic area and in El Placer (21%), the most northerly of the southern reefs. After pooling the data for each subreef, the southern area was considerably less affected overall than the central area (mean = 7.1 %, sd = 9 versus mean = 19.4%, sd = 1 1.7, respectively; for each n = 12 subreefs). White plague (WPD) was the only disease recorded in Boca Paila, where M faveolata constituted 50-75% of all diseased colonies. WPD was responsible for 66-88% of all diseased colonies in Tampalam (M annularis species complex » A. agaricites and P. astreoides) and 62-80 % in El Placer (M annularis species complex » S. siderea, A. agaricites, D. strigosa, A. tenuifolia and Meandrina meandrites). Some colonies of M. meandrites at Tampalam and El Placer were affected by black-band disease and by what appeared to be a form of red-band disease. Dark-spots disease (in S. siderea) and unknown diseases were also seen at Tampalam and a colony of Diploria strigosa with signs of a disease resembling yellow-blotch disease was seen at El Placer. Algae Turf algae were the predominant algal group at all but one of the most northerly subreefs (Table 3). The mean relative abundance of macroalgae was higher in the central area and in El Placer (20-41%, n = 15 subreefs) than further south (13-19%, n = 9 subreefs). Crustose coralline algae were more abundant than macroalgae in most (8/9) subreefs in the three most southerly reefs and they were comparable to macroalgae in relative abundance in El Placer and in the central area. Mean macroalgal canopy heights showed comparatively little variation among reefs, ranging between 3 cm in Xcalak and -5.5 cm in Tampalam. Mean macroalgal indices (macroalgal relative abundance x canopy height, a proxy for macroalgal biomass) were lowest (<80) in most (seven/nine) of the subreefs at Xcalak, Xahuayxol and Mahahual; the highest values (>110) were found in all subreefs in Punta Allen and Tampalam, in two subreefs at Boca Paila and in El Placer N. Overall, the average macroalgal index was somewhat greater in the central area (mean = 1 14.8, sd = 61, n = 838 quadrats) than in the southern area (mean = 82.0, sd = 60 cm, n = 875 quadrats). Relationships Significant relationships detected via correlation and regression analyses include: ( 1 ) positive relationships between mean coral maximum diameter and live coral cover (r = 0.60, r = 0.36, p <0.01); (2) positive relationships between mean density of stony coral recruits and (a) mean live coral cover (r = 0.50, r^ = 25.4, p = 0.01) and (b) mean macroalgal relative abundance (r = 0.45, r = 20.8, p = 0.025); (3) positive relationships between standing dead colonies and nearest human population size (r = 0.58, r^ = 33.4, p <0.01); and (4) inverse relationships between mean macroalgal abundance or macroalgal index and mean herbivorous fish density by subreef (r = -0.44, r^ = 19.5, p = 0.03; r = - 0.55, r^ = 30.3, p <0.01, respectively; fish data from Nunez-Lara et al, this volume). No significant (p >0.05) relationships were found among the other tested variables. 329 DISCUSSION The species composition of the predominant large (>25 cm in diameter) stony corals was relatively homogeneous among the subreefs of any given reef. Montastraea faveolata was the most frequently encountered species at 9/12 subreefs in the central area, whereas in the southern area M. annularis was the most abundant species at over half (7/12) of the subreefs. As M faveolata grows larger than M annularis (Weil and Knowlton, 1 994), it is not surprising that maximum diameters were slightly larger in the central area reefs even though total cover overall was the same in both areas. The high rates of recent partial-colony mortality in the central area and at El Placer (the most northerly reef in the southern area and, like Tampalam, in the Sian Ka’an subregion) are perhaps explained by the high percentage of diseased corals that were present at the time of our surveys (~9.5-30. %, n = 5 reefs). Delayed effects of the 1998 mass bleaching event may have been partially responsible for the high rates of recent mortality in the three most southerly reefs (Mahahual, Xahuayxol, Xcalak) where relatively few corals showed signs of disease. Coral bleaching can be manifested in different intensities, can occur locally or over large geographic areas, and can have diverse causes. Mass bleaching events are generally associated with periods in which sea surface temperature (SST) is above average (at least 1°C higher than the summer maximum, Hoegh-Guldberg, 1999). The scleractinians that were bleached during this study were localized in two of the three reefs in the southern area that were surveyed in June-July, 1999 (Mahahual and Xahuayxol; Table 2). During April-July of 1999, SST averaged ~29.3°C in the southern area of Quintana Roo (SST/AVHRR images courtesy of Ocean Remote Sensing Group, Johns Hopkins Applied Physics Laboratory, Garza-Perez et al., in preparation). We suspect that a minor warming which continued over several months may have triggered the bleaching observed in these two reefs and/or perhaps the stony corals had not yet fully recovered from the effects of the mass bleaching event in 1998. Damage from the passage of Hurricane Mitch in October 1998 is also possible, although unlikely to have contributed very much to total mortality given the relatively high values for large standing dead colonies of Acropora palmata in Xcalak, which is probably more exposed to storm waves than Mahahual and Xahuayxol located in the lee of Banco Chinchorro (Fig. 1). Acroporids had earlier experienced massive mortality around the Caribbean (Goreau et al., 1998), primarily from white-band disease which presumably had also killed the standing dead colonies of A. palmata in Xcalak, Punta Allen and, to a lesser extent, in Punta Yuyum. The demise of Diploria at Punta Yuyum may either have resulted from some combination of bleaching-related mortality {Diploria labyrinthiformis was one of the most affected species during the 1998 mass bleaching event, Garza-Perez, 1999) or from mortality produced by disease. It was anticipated that the reefs in Mahahual and Xcalak (southern area) and in Punta Allen (central area) might have the lowest live stony coral cover and/or the highest rates of disease or mortality, given their proximity to long-established villages (~ 50 years; Bastida-Zavala et al., 2000). In fact, standing dead colonies presented the only significant positive relationship with human population size. Even this relationship could be compromised if colonies that we were unable to identify as standing dead were actually present in higher abundances in the other reefs and/or if Acropora palmata 330 happened to have been relatively more abundant historically at 10 m in the Xcalak and Punta Allen fore reefs than elsewhere. Nevertheless, in terms of live coral cover, old mortality, and standing dead corals, Xcalak, which is one of the oldest settlements in the Mexican Caribbean, was the most impacted of the eight study reefs. Our results are in agreement with Garza-Perez’s (1999) comparative analysis of the same spur-and-groove habitat in Xcalak, Mahahual, Tampalam, and Akumal. Xcalak was so important in fishing and coconut cultivation at the beginning of the 20th Century that an artificial canal was constructed to connect the fisherman of the coast with the city of Chetumal (the present capital of Quintana Roo). Long-term anthropogenic impacts, such as wastewater discharge and fishing, could have contributed to its current condition. However, Xcalak was also the closest study reef to the prolonged high SST anomaly that resulted in high rates of bleaching-related mortality in Belize during the 1998 ENSO event (see Peckol et al., this volume). Herbivorous fishes and echinoids are important determinants of the distribution and abundance of benthic algae and stony coral communities in the wider Caribbean (Wood, 1999). Increases in the abundance of macroalgae are correlated with the mass mortality of Diadema antillarum throughout the region in 1983-84 and with declines in the abundance of herbivorous fishes (Porter and Meier, 1992; Steneck, 1994). The converse has also been found when macroalgae decline after D. antillarum reappears (Aronson and Precht, 2000). Given the almost total absence of D. antillarum in our surveys, and the inverse relationships between mean herbivorous fish density (Nunez- Lara et al., this volume) and both the mean relative abundance of macroalgae and macroalgal index, differential fishing inside and outside of Sian Ka’an Biosphere Reserve may have exerted a primary control on the abundance and biomass of macroalgae in the central-southern Quintana Roo reefs. In other words, the somewhat higher relative abundance of macroalgae in the central area compared to the southern area may be a consequence of reductions in the local herbivorous fish community as populations of piscivorous fishes have rebounded within the reserve. Effects of protection on the abundance of benthic organisms and fishes have been demonstrated previously in the Mexican Caribbean (Arias-Gonzalez, 1998; Arias-Gonzalez et al., 1999) and other parts of the world (e.g., Edgar and Barrett, 1999). If the presence of coral recruits is a sign that a reef is maintaining itself, and has the potential to continue to do so in the future, then all the reefs in the study area are in good condition. However, the complete lack of correspondence between genera that are most abundant as large colonies and those present as recruits (Fig. 2 versus Fig. 4) may suggest a dominance of asexual reproduction in the maintenance of the primary reef- building corals, as has been reported elsewhere in the Caribbean (Jackson, 1985). Hence, the 50% decrease in live stony coral cover that had occurred throughout the region between the early 1990’s and 1999 (from ~25% to -12%) and the high rates of recent partial-colony mortality (7-27.5%) seen in our 1999 surveys are cause for concern regarding the future status of these reef ecosystems. Further indications of widespread disturbance are provided by the size-frequency distributions of most species of large corals being skewed towards small-intermediate sizes (Fig. 3A, B). In general, it can be said that the condition of the stony corals at 10 m depth in the fore-reef spur-and-groove habitat in the central and southern areas of the Mexican Caribbean already has deteriorated from natural and possibly anthropogenic 331 perturbations, particularly the effects of diseases (in stony corals and Diadema) and bleaching. Its fish communities are also being affected by extractive activities, particularly in the southern area (see Nunez-Lara et ah, this volume), possibly leading to further deterioration in total reef condition. Thus the introduction of new disturbances related to urbanization and tourist development in the southern area (Daltabuit Godas, 1999) could potentially overwhelm the natural resilience of these coastal coral reefs (i.e., Nystrom et ah, 2000). ACKNOWLEDGMENTS This project was financed by the Caribbean Enviromuent Programme of the United Nations Environment Programme, Consejo Nacional de Ciencia y Tecnologia (CONACYT) and CINVESTAV. The AGRRA Organizing Committee also facilitated financial support as described in the Forward to this volume. We express our sincere thanks to the Secretaria de Medio Ambiente, Recursos Naturales y Pesca (SEMARNAP) in Quintana Roo for allowing us to use their facilities at Mahahuah Thanks are also due to Biologist Alfredo Arrellano Guillermo, Director, Sian Ka’an Biosphere Reserve (RBSK) for his invaluable logistical assistance during the work within the Reserve; to Amigos de Sian Ka’an for permission to use Pez Maya as a work base within the Reserve; and to Chente, Chepe, Felipe, Fredy and Vidal for their valuable help during the fieldwork. We especially thank Judy Lang and Robert Ginsburg for comments and suggestions on the first version of the manuscript. REFERENCES Arias-Gonzalez, J.E. 1998. Trophic models of protected and unprotected coral reef ecosystems in the south of the Mexican Caribbean. Journal of Fish Biology 53 (Supplement A): 236-255. Arias-Gonzalez, J.E., V. Cataneda, C. Gonzalez-Salas, C. Caceres, R. Garza, A. Maldonado, and E. Nunez-Lara 1999. What effect can minimum restrictions produce in a coral reef area? Coral Reefs 18:212. Aronson R.B., and W.F. Precht 2000. Herbivory and algal dynamics on the coral reef at Discovery Bay, Jamaica. Limnology and Oceanography 45(1):25 1-255. Bastida-Zavala, J.R., A.U. Beltran-Torres, M.A. Gutierrez-Aguirre, and G. de la Fuente- Betancourt 2000. Evaluacion rapida de los arrecifes parche de Majagual, Quintana Roo, Mexico. Revista de Biologla Tropical 48(1): 137-143. 332 Beltran-Torres, A.U., and J.P. Carricart-Ganivet 1999. Lista revisada y clave para los corales petreos zooxantelados (Hydrozoa: Milleporina; Anthozoa: Scleractinia) del Atlantico Mexicano. Revista de Biologia Tropical 47(4):8 13-829. Daltabuit Godas, M. 1999. Desarrollo turistico en Quintana Roo y sus efectos en el sistema arrecifal mesoamericano y del Caribe. 22 Pp. In: IX Reunion Anual del Programa Universitario del Medio Ambiente, 1999: Arrecifes Coralinos, Riqueza Marina Amenazada. UNAM, SEMARNAP, PROFEPA. CD-ROM ISBN 968-36-8434- 3. Edgar, G.J., and N.S. Barrett 1999. Effects of the declaration of marine reserves on Tasmanian reef fishes, invertebrates and plants. Journal of Experimental Marine Biology and Ecology 242:107-144. Garza-Perez, J.R. 1999. Andlisis Comparativo de Cuatro Comunidades Coralinas Arrecifales del Caribe Mexicano. Master Degree Thesis. Centro de Investigacidn y Estudios Avanzados del I.P.N., Unidad Merida, Mexico. 64 pp. Garza-Perez, J.R., L.O. Santos-Rodriguez, and J.E. Arias-Gonzalez 2000. Bleaching and hurricane impact assessment in Mahahual Reef, Mexico. Ninth International Coral Reef Symposium, Bali. p. 374 (Abstract). Goreau T. J., J. Cervino, M. Goreau, R. Hayes, M. Hayes, L. Richardson, G. Smith, K. DeMeyer, I. Nagelkerken, J. Garzon-Ferrera, D. Gil, E.C. Peters, G. Garrison, E.H. Williams, L. Bunkley-Williams, C. Quirolo, and K. Patterson 1998. Rapid spread of diseases in Caribbean coral reefs. Revista de Biologia Tropical 46(Supplement 5): 157- 172. Gutierrez Carbonell D., G. Garcia Beltran, M. Lara Perez Soto, C. Padilla Souza, J. Pizana Alonso, and O. R. Macias 1993. Caracterizacion de los arrecifes coralinos de la Reserva de la Biosfera de Sian Ka’an, Q. Roo. Sian Ka ’an Serie de Documentos 1 : 1-47. Gutierrez Carbonell D., M. Lara Perez Soto, C. Padilla Souza, J. Pizana Alonso, G. Garcia Beltran R., M. Loreto Viruel, and T. Camarena Luhrs 1995. Caracterizacion de los arrecifes coralinos en el corredor “Cancun-Tulum,” Quintana Roo, Mexico. Sian Ka ’an Serie de Documentos 4:3-39. Hoegh-Guldberg, O. 1999. Climate change, coral bleaching and the future of the world’s coral reefs. Marine and Freshwater Research 50:839-66. Humann, P. 1993a. Reef Coral Identification, Florida, Caribbean, Bahamas. New World Publications, Inc., Jacksonville, FL. 239 pp. Humann, P. 1993b. Reef Creaturel Identification, Florida, Caribbean, Bahamas. New World Publications, Inc., Jacksonville, FL. 320 pp. 333 ICRI 1998. The International Coral Reef Initiative: The Status of Coral Reefs in Mexico and the United States Gulf of Mexico (geography, ecology, management, legislation, monitoring, and bibliography) .CD-ROM.. U.S. Department of Commerce, National Oceanic and Atmospheric Administration; The Commission for Environmental Cooperation; Amigos de Sian Ka’an; Centro de Investigacion y de Estudios Avanzados del Instituto Politecnico Nacional- Unidad Merida; The Nature Conservancy. INEGI 1996. Quintana Roo conteo de poblaciony vivienda 1995, resultados definitivos tabulados bdsicos. Instituto Nacional de Estadistica Geografia e Informatica. Mexico. 219 pp. INEGI 2000. Sistema de Cuentas Nacionales de Mexico. Producto Intemo Bruto por Entidad Federativa, 1993-1999. Instituto Nacional de Estadistica Geografia e Informatica. Mexico. 522 pp. Jackson, J.B.C. 1985. Distribution and ecology of clonal and aclonal benthic invertebrates. Pp: 297-355. In: J.B.C. Jackson, L.W. Buss, and R.E. Cook (eds.). Population Biology and Evolution of Clonal Organisms. Yale University Press. New Haven. Jordan, D.E. 1993. Atlas de los Arrecifes Coralinos del Caribe Mexicano. Parte I El Sistema Continental. Centro de Investigaciones de Quintana Roo, Instituto de Ciencias del Mar y Limnologia. 1 10 pp. Nystrom M., C. Folke, and F. Moberg 2000. Coral reef disturbance and resilience in a human-dominated environment. TREE 15:413-417. Nunez-Lara, E., and J.E. Arias-Gonzalez 1998. The relationship between reef fish community structure and environmental variables in the southern Mexican Caribbean. Journal of Fisheries Biology 53(A): 209-221. Oxley, W.G. 1997. Sampling design and monitoring. Pp. 307-320. In: S. English, C. Wilkinson and V. Baker (eds.). Survey Manual for Tropical Marine Resources. Australian Institute of Marine Science, Townsville, Australia. Porter J.W., and O.W. Meier 1 992. Quantification of loss and change in Floridian reef coral populations. American Zoologist 32:625-640. 334 Steneck R.S. 1994. Is herbivore loss more damaging to reefs than hurricanes? Case studies from two Caribbean reef systems (1978-1988). Pp. 220-226. In: R.N. Ginsburg (compiler), Global Aspects of Coral Reefs - Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Weil, E., and N. Knowlton 1994. A multi-character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bulletin of Marine Science 55:151-175. Table 1. Site information for AGRRA stony coral and algae surveys in central-southern Quintana Roo, Mexico. 335 Table 2. Size and condition (mean ± standard deviation) of all stony corals (>25 cm diameter) by subreefs in central-southern Quintana Roo, Mexico. 336 >, P2 c o CO C/5 S.O ^ 3 ^ on ^ x: (D O > 3 •Jj ^ X) 00 (N ON ^ so CsJ — . ^ ^ ^ o o o O O O (N O O o o o in w-t ^ ^ -I D E 3 E 3 E 3 2: c CJ c GO c o z D u o CO r, o c flj 3 o o iz; C CJ U 3 O CO o Z D U O CO Cu ’c5 o- jd cd O- 3 >> 3 >% 3 < — < < E 2 cd E 2 cd Id z; U| (U CJ CJ (D -a 03 ^3 c 03 +1 c o3 (U £ s: <3 Q TJ C 03 O . c o ■(— * t/3 o o _o - _c 03 '£ 03 a O E 13 w) < =s o , 1/3 r<3 1 3 3 c3 H ■t— > c (U o 338 Figure 1. AGRRA survey reefs in central-southern Quintana Roo, Mexico. Modified from Niinez-Lara and Arias-Gonzalez (1998). CONDITION OF CORAL REEF ECOSYSTEMS IN CENTRAL-SOUTHERN QUINTANA ROO (PART 2: REEE EISH COMMUNITIES) BY ENRIQUE NUNEZ-LARA,‘ CARLOS GONZAlEZ-SALAS/’^ MIGUEL A. RUIZ-ZArATE,' ROBERTO HERNAnDEZ-LANDA,* and J. ERNESTO ARIAS-GONZALEZ' ABSTRACT Increases of fishing and tourism threaten the natural relationships between reef fish communities and their environments. All species of reef fishes were visually assessed in the central and southern Mexican Caribbean in eight fringing reefs, four of which are in a protected biosphere reserve. The sampling design ineluded three spatial scales from tens of meters to tens of kilometers. A total of 9,908 individuals belonging to 128 species and 43 families were identified in 144 belt transects. Zooplankton feeders were the most important trophic group by number of individuals; plant and detritus feeders dominated by number of species. Herbivores were larger in unprotected reefs than in the reserve. Regression analyses showed significant inverse relationships between total fish species density and macroalgal index (a proxy for macroalgal biomass) and, for “large” (>25 cm diameter) stony corals, partial-colony mortality and live/dead ratio. Significant inverse relationships were also found between mean abundance of the plant and detritus feeders guild and macroalgal index and macroalgae abundance. Geomorphological factors and anthropogenic impacts, both positive (protection) in the reserve and negative (fishing and tourism) in unprotected areas, may explain these spatial patterns in reef-fish community structure. INTRODUCTION The reefs of the Mexiean Caribbean run along the eastern margin of the Yucatan Peninsula. They are distributed parallel to the coastline of Quintana Roo state as a fringing reef system which originated during Miocene-Pleistocene rifting of the carbonate platform (Weidie, 1985). Eight reefs with similar structural configurations were ehosen for the present study (Fig. 1). The reef profile can generally be divided into three main ' Laboratorio de Ecologia de Ecosistemas Arrecifes Coralinos, Centro de Investigaeion y de Estudios Avanzados IPN, Carretera Antigua a Progreso Km 6, AP 73, Merida, Yucatan, Mexico. Emails: enunez@mda.cinvestav.mx, earias@mda.cinvestav.mx 2 Ecole Pratique des Hautes Etudes, Laboratoire dTchtyoecologie Tropicale et Mediterraneenne, ESA 8046 CNRS, Universite de Perpignan, 66860 Perpignan cedex FRANCE. E-mail: cgonzale@univ-perp.fr 340 zones: reef lagoon, reef crest, and fore reef (mostly spur-and-groove). Four of the reefs are located within the Sian Ka’an Biosphere Reserve, which extends from 19°05’N to 20° 06’N. The reserve, which was created in 1986 by presidential decree, has approximately 120,000 ha of coastal environments including 37,000 ha of coral reefs that were added in 1998. Natural ecological conditions within the reserve have been conserved due to the long distance from large population centers, limited access, and the restrictions on fishing and tourism. Punta Allen (Rojo Gomez) and Punta Herrero are the reserve’s two largest human communities, both having less than 300 inhabitants. The villagers are principally dedicated to lobster fishing and less effort is given to catching bony fishes of which the main species are mojarras (Gerres spp.), snappers (lutjanids), barracudas (Sphyraena barracuda), grunts (haemulids), and groupers (serranids). The remaining reefs are located between the southern boundary of the Sian Ka’an Biosphere Reserve and the frontier with Belize. The predominant activity in this area is fishing, mainly for local consumption. Most fish are caught in Banco Chinchorro, a shelf- edge bank reef system located about 45 km seaward of Mahahual town. The principal fishing methods employed are trotlines, traps, gill nets and harpoons [Oficina Regional de Pesca (SEMARNAP), unpublished report]. The Regional Fishing Office for 1997 reported a total fish catch of 335,443 kg from the zone between Punta Herrero, in the southern of Sian Ka’an reserve, and Xcalak, near the Belize border. However, locality- specific information was not provided. Tourism has recently begun to grow rapidly along the unprotected southern coast of the Mexican Caribbean. The reefs under greatest threat, Mahahual and Xcalak, are in the path of large developers. Tourism potentially constitutes a relatively benign and lucrative use of coral reef resources. However, this benefit can be counteracted by damage and overexploitation (Hawkins and Roberts, 1993). The relationship between tourism and reef fishes is mainly indirect, being felt through the effects on fish habitat (Russ, 1991) including coral breakage and death due to vessel anchors, sedimentation, dredging, and other coastal zone activities (Dollar, 1982; Grigg, 1994; Muthiga and McClanahan, 1997). A direct effect of tourism is the extraction of fishes for consumption by visitors. Changes in community structure are caused by overexploitation of fishes of the high trophic levels (Russ, 1991). Thus, effects of fishing are partially attributable to the high demand of fishes for tourists and partially to the continuous increase in the numbers of fishermen and local inhabitants. There are relatively few studies of reef fish communities in the Mexican Caribbean (e.g., Fenner, 1991; Diaz-Ruiz and Aguirre-Leon, 1993; Schmitter-Soto, 1995; Arias-Gonzalez, 1998; Diaz-Ruiz et al., 1998; Nunez-Lara and Arias-Gonzalez, 1998). In the present study we tried to detect the main structural forces affecting reef fish communities, and to measure the influence of fisheries and tourism on fishes and their habitat. The data generated represents a basis for comparative analyses of reef fish community structure at different spatial scales. METHODS Reef fishes were visually censused at eight fringing reef localities along Mexico’s central and southern Caribbean coast. The reefs were selected on the basis of a mixture of 341 strategic and representative criteria, including their spatial separation distance (usually 20- 30 km), natural geographic barriers (such as the two large bays that divide the Sian Ka’an Biosphere Reserve), and their type of use by humans. For the purpose of this study, every reef was assigned to one of three geographical areas (Fig. 1): Northern Sian Ka’an (NSK) at 19°-20° N (3 reefs about 25 km apart); Southern Sian Ka’an (SSK) at 18°-19° N (2 reefs about 30 km apart-one being outside the reserve); and Southern (S) at 17°-18° N (3 reefs about 30 km apart). The distance parallel to the reef crest that constituted each reef was approximately three kilometers. This distance was subdivided into three 0.9-1 km subreefs (north, center, and south). Subreefs were geographically localized with GPS and described in terms of distance from the coast and degree of exposure to oceanic currents. Six replicate belt transects, each measuring 50 m long by 2 m wide, were swum parallel to the coast at every subreef. Transects were spaced approximately 100 m apart and all surveys were made between 0900 and 1700 hours by one diver (Nuilez-Lara). All transects were made at an average depth of 12 meters in the fore reef. The dominant habitat was spur and groove, except in northern Tampalam, where the calcareous substratum was largely covered with benthic algae, gorgonians, and sponges. All reef fishes >3cm in body length within the transects were counted and their sizes estimated in six categories: 3-10 cm, 11-20 cm, 21-30 cm, 31-40 cm, 41-50 cm and > 50 cm. The Atlantic and Gulf Rapid Reef Assessment (AGRRA) fishes constitute a subset of the “all species” data: in this paper, “serranids” are species of Epinephelus (including E. fulvus but excluding E. cruentatus, which is here classified as Cephalopholis cruentata) and Mycteroperca; “haemulids” and “scarids” (parrotfishes) refer to fishes that are >3 cm in total length. In order to describe the reef fish community structure, the following ecological descriptors were calculated: species richness, abundance, density, trophic structure and size structure. Three different spatial scales were used for the analysis: subreef (hundreds of meters), reef (kilometers) and area (tens of kilometers). Trophic structure was analyzed by calculating the percentage of individuals and fish species belonging to each of Randall’s (1967) feeding categories: plant and detritus feeders; zooplankton feeders; sessile invertebrate feeders; “shelled” invertebrate feeders; generalized carnivores; ectoparasite feeders; and fish feeders. A multiple regression technique was used to relate the total fish density with the following benthic habitat variables assessed by Ruiz et al. (this volume): total live stony coral cover; total (recent + old) partial-colony mortality, old partial-colony mortality, recent partial-colony mortality, live:dead ratio and maximum diameter for “large” (>25 cm in diameter) stony corals; and relative abundance and macroalgal index (relative abundance x height, a proxy for biomass) for macroalgae. Simple regression was used to examine the relationship between total fish herbivore density and the two macroalgal descriptors (index and relative abundance). Densities were log+1 transformed to meet the assumptions for parametric regression tests. Classification analysis was performed to determine the degree of similarity among subreefs, based on the abundance values of their recorded fish species. The Bray-Curtis (1957) distance index was used as a similarity measure and the Unweighted Pair Grouping Method Average (UPGMA) as a clustering method. 342 Figure 2. Species richness (all fishes >3 cm long) by subreef at 12 m depth in central-southern Quintana Roo, Mexico. 343 Figure 3. Mean density (no. individuals/ 100 m^) of all fishes (>3 cm long), and of AGRRA fishes, by subreef at 12 m depth in central-southern Quintana Roo, Mexico. 344 RESULTS Species Composition and Richness A total of 128 reef fish species belonging to 43 families were identified in 144 belt transects. The greatest number was found in the NSK area and the lowest in the S area (Table 1, Fig. 2). Particularly notable was the large number of species belonging to the families Holocentridae (squirrelfish), Serranidae (grouper) and Haemulidae (grunts) in the NSK area and the relatively low number of species from such “typical” reef fish families as Pomacentridae (damselfish), Labridae (wrasse), and Scaridae (parrotfishes) in the SSK area. A somewhat different pattern was evident for AGRRA fishes (Table 1) as species numbers overall were rather similar in the NSK and S areas (23-32, n = 18 subreefs). In the SSK area, the number of AGRRA species was slightly lower at El Placer (21-27, n = 3 subreefs), but much smaller at Tampalam (n = 5-15, n = 3 subreefs). The 25 dominant fish species in terms of sighting frequency and density belonged to the following families: Labridae, Acanthuridae (surgeonfish), Scaridae, Pomacentridae, Haemulidae, Serranidae, Lutjanidae (snappers), Pomacanthidae (angelfish), Holocentridae and Grammatidae (basslet). Thalassoma bifasciatum was the most frequently sighted and abundant species in all the studied reefs, followed by Acanthurus coeruleus, Sparisoma aurofrenatum and Halichoeres garnoti (Table 2). Forty percent of these dominant fish species are included in the AGRRA fish list. Abundance and Density A total of 9,908 fishes were counted in the 144 transects, 3,733 of which belonged to the species on the AGRRA list. Total densities were highest in Xcalak S and Punta Yuyum C and lowest in Tampalam N and Xahuayxol C (Fig. 3). In an examination of the key AGRRA families, regardless of scoring mode (all individuals or restricted counts for haemulids, scarids, and serranids), the Scaridae was found to have the greatest density. 50 Fish families Figure 4. Mean density (no. individuals/lOOm"^ ± se) of all fishes (>3 cm long), and of AGRRA fishes, by family in central-southern Quintana Roo, Mexico. Other AGRRA fishes = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. 345 followed by the Haemulidae and Acanthuridae (Fig. 4). Although less abundant, the Lutjanidae and Serranidae were more plentiful than the Balistidae (leather] ackets), Chaetodontidae (butterflyfish), and Pomacanthidae. Parrotfishes >3cm and surgeonfishes were most abundant overall in the S area, although the density of parrotfishes was also high at Punta Allen (in the NSK area). Surgeonfishes were relatively scarce at Mahahual (in the S area) where scarid density was highest (Table 3). The density of snappers was highest in Boca Paila and Punta Yuyum reefs in the NSK area and in Xcalak in the S area. Grunts >3cm were also abundant in Punta Yuyum (especially) and Boca Paila as well as in El Placer (SSK area). The density of groupers {Epinephelus, Mycteroperca) was low in all the reefs. Trophic Structure Considering total abundances, the trophic structure of the fish community was dominated by zooplankton feeders (39%), followed by plant and detritus feeders (28%) and shelled invertebrate feeders (16%). In term of total species richness, the three most important of Randall’s (1967) trophic groups were the plant and detritus feeders (24%), “shelled” invertebrate feeders (21%), generalized carnivores (15%) and sessile invertebrate feeders (14%). Ectoparasite feeders and fish feeders showed the lowest percent contribution for both individuals and species (Fig. 5). Similar patterns were detected for every area and each of the sampled reefs, with zooplankton feeders dominating by number of individuals and the plant and detritus feeders being the most important group by numbers of species, closely seconded by “shelled” invertebrate feeders (Table 4). Among the most abundant of the herbivorous fish species were Scarus iserti (=S. croicensis), Acanthurus coeruleus, A. bahianus, Sparisoma aurofrenatum and S. viride. Figure 5. Trophic structure (as percent of species and individuals) for all fishes >3 cm long at 12 m in central-southern Quintana Roo, Mexico. 346 I B 0^ l*3-10cm ihn-20cm 111=2 lOOcm IV=31-40cm V=>40cm Figure 6. Size frequency distribution of (A) all plant and detritus feeders >3 cm (acanthurids, kiphosids, pomacentrids except Chromis, and scarids) and AGRRA herbivores >3 cm (acanthurids, scarids, Microspathodon chrysurus) and (B) all carnivores >3 cm [carangids, lutjanids, scombrids, select serranids {Epinephelus and Mycteroperca), sphyraenids] and AGRRA carnivores >3 cm [lutjanids, Epinephelus (except for E. cruentatus) and Mycteoperca] at 12 m in central-southern Quintana Roo, Mexico. Lengths Very few of the surveyed fishes exeeeded 30 cm in length. The most common size classes for “all herbivores” (acanthurids, kyphosids, pomacentrids except Chromis, and scarids) were equally divided between the 3-10 cm and 11-20 cm length intervals (Fig. 6A). In a more detailed analysis, we observed proportionately more fishes in the 11- 20cm size class in the S area (particularly in Mahahual and Xcalak) than in the NSK and SSK areas where smaller herbivores (3- 10cm) were relatively more abundant (Table 5). Key herbivores (acanthurids, scarids >3cm, Microspathodon chrysurus) were slightly larger (10-20 cm) overall (Fig. 6A) and also attained their largest sizes in the S area. Both for “all carnivores” (carangids, lutjanids, scombrids, sphyraenids, plus Epinephelus and Mycteroperca) and for the AGRRA carnivores (lutjanids, select serranids), the most common size class was 1 1-20 cm (Fig. 6B). Among the more abundant carnivores (total and AGRRA species) were the relatively small-sized Epinephelus fulvus, Lutjanus apodus and L. synagris. However, carnivores were slightly larger in the reefs in which they were most abundant (Xcalak, Boca Paila and Yuyum) (Table 5). 347 Classification Analysis Classification analysis divided the sites into two large clusters according to the affinity criteria based on total fish abundances. The first cluster of 10 subreefs includes only sites located in the NSK and SSK areas. The second cluster includes all nine of the subreefs of the S area, three of the NSK area and El Placer S subreef (SSK area). Tampalam N subreef was markedly different from all the other survey sites (Fig. 7). Figure 7. Hierarchical classification analysis based on total fish abundance, by subreef at 12 m in central- southern Quintana Roo, Mexico. Relationships Multiple regression analyses between the total density of all the fish species and the benthic habitat variables (Ruiz et al., this volume) showed statistically significant inverse relationships (P < 0.05) with the macroalgal index, live/dead stony coral ratio, and total (recent + old) partial-colony mortality (Fig. 8). The r^ statistic indicates that the model of multiple regression, when all six benthic variables are included, explains 86.96% of the variability in the fish density data. Statistically significant, inverse relationships were also found between the means for the total plant and detritus guild and those for the macroalgal index and relative abundance of macroalgae (Fig. 9). 348 25 cm in diameter)stony corals (B) mean live:dead ratio (P<0.05, 26.3% explained), (C) mean total (recent + old) partial colony mortality (P<0.05, 22.5% explained) by subreef at 12 m in central-southern Quintana Roo, Mexico. 349 o o % T3 u Ui O > X Plant and detritus feeders density = 31.9796-0.12449 * macroalgal index 40 60 80 100 120 140 160 A. Macroalgal index Plant and detritus feeders density = 31.4692 - 0.563723 * macroalgal relative abundance o o % c 3 cm) Haemulidae (>3 cm) Lutjanidae Serranidae' Northern Sian Ka ’an Boca Paila N 4.7 ±2.4 5.8 ±3.5 3.5 ± 1.5 2.7 ±3.1 2.0 ±0.8 Boca Paila C 8.0 ±7.3 2.0 ± 1.5 11.7± 8.1 3.0 ±3.3 0.7 ± 1.0 Boca Paila S 7.7 ±6.8 3.8± 1.8 13.3 ± 10.4 6.8 ± 1.2 1.2 ±2.5 Yuyum N 7.5 ±2.0 8.0 ±5.5 12.0 ±7.6 2.5 ±2.6 1.5 ±2.0 Yuyum C 4.7 ±3.7 8.2 ±2.5 24.3 ± 14.9 9.5 ± 7.6 2.0 ±2.5 Yuyum S 4.8 ± 5.5 11.0 ±4.0 7.3 ±4.1 1.2 ±0.8 1.2 ±0.5 Punta Allen N 3.2 ±5.1 9.8 ±5.7 2.0 ± 1.2 1.3 ±2.2 1.8 ±0.8 Punta Allen C 4.2 ± 1.5 9.8 ±5.7 3.0 ± 1.9 1.0 ±0.6 2.3 ±2.0 Punta Allen S 3.5 ±2.5 15.7± 12.1 6.3 ±8.7 1.3± 1.5 1.3 ±0.8 Southern Sian Ka ’an Tampalam N 5.0 ±4.0 2.0 ±2.0 5.0 ±0.5 0 1.7 ±4.0 Tampalam C 4.7 ±9.7 6. 5 ±5.0 7.0 ±4.4 2.5 ± 1.3 0.3 ±0.8 Tampalam S 4.0 ±8.0 10.8 ±5.9 9.0 ±6.7 2.7 ±2.2 0.2 ±0 El Placer N 3.8 ±4.2 5.2 ±2.9 15.5 ± 18.7 0.7 ±0.5 2.3 ±4.2 El Placer C 1.8 ±0.8 9.0 ±5.1 4.0 ±2.5 0.2 ± 0.6 1.0 ±2.0 El Placer S 3.8 ±3.7 12.2 ±7.4 6.3 ± 5.7 1.8± 1.5 1.8 ±3.5 Southern Mahahual N 2.5 ±0.5 15.0 ± 12.0 1.0± 1.3 1.2±0 1.5± 1.3 Mahahual C 2.5 ±3.5 21.8 ± 17.8 2.3 ± 1.5 0.3 ±0 0.8 ±0.8 Mahahual S 3.3 ±3.5 25.0 ± 13.4 2.3 ± 1.0 1.0 ±0.6 1.5 ±0.9 Xahuayxol N 6.7 ±7.7 13.0 ±4.7 2.8 ± 1.3 1.0 ± 10.3 1.8 ±0.6 Xahuayxol C 4.5 ±4.6 8.3 ±4.8 2.5 ±0.8 1.0 ±0.6 2.8 ±6.5 Xahuayxol S 6.2 ±4.8 11.3 ±6.7 7.5 ±6.0 2.0 ±3.0 2.3 ±3.5 Xcalak N 4.0 ±4.0 13.2± 5.1 2.7 ±0.6 6.7 ± 12.4 1.7± 1.0 Xcalak C 5.0 ±3.3 10.0 ±4.3 4.3 ±3.0 2.7 ±3.5) 1.2 ±2.5 Xcalak S 16.7 ±22.4 14.2 ±6.4 3.0± 1.1 4.7 ±6.0 2.3 ±2.7 'Epinephelus spp. (excluding any E. cruentatus) and Mycteroperca spp. Table 4. Mean percentage of species and individuals in Randall’s (1967) reef fish feeding categories by reef in central-southern Quintana Roo, Mexico. 357 358 Table 5. Mean abundance by size category for herbivorous and carnivorous reef fishes (all species and AGRRA list) by reef in central-southern Quintana Roo, Mexico. Reef name Fish trophic category Abundance (#/reef) 3-10 cm 1 1-20 cm 21-30 cm 3 1 -40 cm 41-50 cm Northern Sian Ka’an Boca Paila All herbivores 219 132 28 8 7 AGRRA herbivores' 35 128 28 4 4 All carnivores 2 70 29 12 6 AGRRA carnivores' 1 52 12 8 3 Yuyum All herbivores 195 163 33 7 0 AGRRA herbivores 76 163 33 7 0 All carnivores 9 86 27 6 1 AGRRA carnivores 3 65 10 1 0 Punta Allen All herbivores 176 188 21 7 2 AGRRA herbivores 92 184 20 7 1 All carnivores 0 25 19 5 5 AGRRA carnivores 0 5 10 5 5 Southern Sian Ka’an Tampalam All herbivores 204 111 14 9 3 AGRRA herbivores 78 107 14 9 1 All carnivores 0 32 16 3 3 AGRRA carnivores 0 21 8 2 0 El Placer All herbivores 285 99 35 7 0 AGRRA herbivores 97 90 35 7 0 All carnivores 0 51 12 6 8 AGRRA carnivores 0 14 0 2 0 Southern Mahahual All herbivores 173 334 65 4 0 AGRRA herbivores 60 306 65 4 0 All carnivores 4 33 11 2 3 AGRRA carnivores 0 9 5 0 1 Xahuayxol All herbivores 146 218 29 9 0 AGRRA herbivores 76 199 29 9 0 All carnivores 6 51 19 1 1 AGRRA carnivores 0 13 10 0 0 Xcalak All herbivores 181 317 35 11 0 AGRRA herbivores 49 316 35 11 0 All carnivores 6 53 50 25 4 AGRRA carnivores 0 26 34 22 3 All Reefs All herbivores 197.4 195.3 32.5 7.8 1.5 AGRRA herbivores 70.4 186.6 32.4 7.3 0.8 All carnivores 3.4 50.1 22.9 7.5 3.9 AGRRA carnivores 0.4 22.8 9.9 4.4 1.3 'See Methods for AGRRA species as defined in this paper. 359 Plate 8A. Stony coral diseases are recorded in the AGRRA benthos protocol on the basis of color and host identity; white plague as shown in Dichocoenia stokesi and other massive corals; white-band and other white-type diseases in acroporids; black-band in many corals; yellow-blotch (yellow-band) in Montasiraea spp. and brain corals; and dark spots {Siderastrea siderea and Stephanocoenia intersepta). (Photo George P. Schmahl) Plate 8B. Diseases presently constitute some of the most severe disturbances on reefs in the wider Caribbean and they are increasing in frequency, intensity and distribution. Signs of disease were present in about 3% (at <5 m) and 5% (at >5 m) of the stony corals ( >25 cm diameter) observed during the AGRRA surveys reported in this volume. Recent mortality in this colony of Montastraea faveolata was caused by black-band disease (arrow). (Photo Kenneth W. Marks) 360 Figure 1. AGRRA survey sites in the Veracruz Reef System. CONDITION OF SELECTED REEF SITES IN THE VERACRUZ REEF SYSTEM (STONY CORALS AND ALGAE) BY GUILLERMO HORTA-PUGA* ABSTRACT Three platform reefs off the city of Veracruz in the Gulf of Mexico were surveyed during July, 1999 with the Atlantic and Gulf Rapid Reef Assessment (AGRRA) benthos protocols at two depth intervals (3-6 m and 9-12 m) in windward fore-reef habitats. Live stony coral cover averaged 17%. Biodiversity was low with 15 taxa of “large” (>25 cm diameter) stony corals of which Montastraea cavernosa was numerically the most abundant (35%). Acroporids were almost completely absent. The condition of the large living corals was good with few signs of disease (none in the individually assessed colonies) and little bleaching. Crustose coralline algae were more abundant overall than turf algae while marcoalgae were scarce. As coral recruitment density was extremely low (-1.2 recruits/m^), the current potential for recovery of these reefs to historical levels of live coral cover seems poor, even though apparently suitable recruitment sites were available in most sites. INTRODUCTION The 20 coral reefs of the Veracruz Reef System (VRS) are located within 22 km of the coast near the Port of Veracruz in the western Gulf of Mexico (Fig. 1). The reefs developed on the continental shelf after the last glacial period some 9,000 to 10,000 years ago (Morelock and Koenig 1967; Kuhlmann 1975). They have thrived in a naturally turbid environment. Visibility can be <1 m during the rainy season (June-October) when high concentrations of suspended particles (eroded materials from the mainland) are transported by several rivers (Antigua, Jamapa, Papaloapan) that discharge nearby. The area is also affected by several cold fronts each winter that decrease seawater temperature and increase surf and turbidity. The VRS reefs have well-developed reef frameworks (Carricart-Ganivet et al., 1993). The windward fore-reef slopes, which face east and north, have the highest stony coral cover and extend to depths of about 30 m (Gutierrez et al., 1993; Vargas-Hernandez et al., 1993). Live stony coral cover is lower, and limited to depths of about 21 m, on the leeward (westward-facing) slopes which, however, have the highest diversity of scleractinians as certain ramose and delicate species only occur in these areas. The reef flats (also called platforms or lagoons) are well-illuminated, shallow (1.0-1. 6 m in depth), and dominated by ' fNVEMAR, UBIPRO, FES-Iztacala, Universidad Nacional Aut6noma de Mexico Av. de los Barrios 1, Los Reyes Iztacala, Tlanepantla, Mexico 54090, Mexico. E-mail: horta@servidor.unam.mx 362 macroalgae, the seagrass, Thalassia testudinum, and some stony corals. The reef erests or rubble ramparts, which are located at the windward margins of the reef flats, are aceumulations of dead coral boulders deposited by strong storm waves and currents (Kiihlmann, 1975; Gutierrez et ah, 1993). As in other remote Atlantic reef areas like the Flower Garden Banks and Bermuda, the diversity of reef-building scleraetinians is relatively depauperate with 30 zooxanthellate species in the VRS in contrast to the 47 found on Mexican Caribbean reefs in the eastern Yucatan Peninsula (Horta-Puga and Carricart-Ganivet, 1993; Beltran-Torres and Carricart- Ganivet, 1999; Ruiz et ah, this volume). Historically, the main speeies have been Hcropora cervicornis, A. palmata, Diploria clivosa, D. strigosa, Siderastrea radians, S. siderea, Montastraea annularis, M. cavernosa, M. faveolata and Colpophyllia natans (Lara et ah, 1992; Gutierrez et ah, 1993). However, both species of Acropora experieneed widespread mortality during the 1970s- 1980s (Jordan, 1992; Tunnell, 1992), possibly at least in part from disease. At present the bottom is dominated by the long-dead skeletons of important genera {Acropora, Montastraea, Colpophyllia, Diploria, Siderastrea) which are still recognizable suggesting high mortality rates in recent decades. Populations of Diadema antillarum were possibly deeimated during the 1983-84 mass mortality event which changed sea-urchin community structure throughout the Western Atlantic. Nowadays, species of Echinometra {E. lucunter and E. viridis) are among the most abundant herbivores in the VRS (see Results). The reefs of the VRS have been considered to be among the most threatened in the wider Caribbean. Anthropogenic impacts in the area are diverse and include oil spills plus other chemical pollutants, overfishing, unrestricted recreational diving, ship groundings, sewage effluents, dredging, reereational diving, and boat anchoring (Tunnell, 1992; Lang et al., 1998). Environmental stressors have increased in magnitude in recent decades. In response to seientific concern, the entire VRS was decreed a marine protected area by the Mexican Government in 1992 (Diario Oficial, 1992). In 2000, the National Commission of Natural Protected Areas designated the staff of the marine park who are now actively developing a management plan as well as performing surveillance and conservation activities. The aim of this study was to assess the current condition of the VRS using the AGRRA benthos protocols. METHODS The VRS is divided into a northern and a southern group by the outlet of the Jamapa River. The northern group has three coastal fringing reefs and seven platform reefs rising 25- 30 m above the bottom (Carricart-Ganivet and Horta-Puga, 1993). Three platform-type mid- shelf reefs (Fig. 1) were ehosen to assess the benthic condition of the northern VRS reef system. Galleguilla was strategically chosen because it is considered highly threatened due to its position near the outlet of Veraeruz City’s sewage treatment plant and the port docks. Isla Verde, which is furtherest offshore, and Isla de Sacrificios are both considered representative platform reefs. The latter is also a strategic site because its reef has been closed to all human aetivities related to tourism or fishing by local authorities since 1982 and was hoped to be in a process of reeovery after having been highly degraded. The surveys were conducted in windward (eastern), fore-reef slope habitats. Two depth zones were selected: the shallow (3-6 363 m) former Acropora zone and a deeper (9-12 m) zone of maximum scleractinian diversity and coverage. The AGRRA Version 2 benthos protocols (see Appendix One, this volume) were followed from 14-22 July 1999 with the following modifications; (1) colonies that were “standing dead” (completely dead and still in growth position) were not surveyed, i.e., individual assessments were restricted to “large” (>25 cm maximum diameter) corals with at least some living tissues; (2) coral measurements were made to the nearest 5 cm; (3) Diploria, Siderastrea and Agaricia were identified only to the genus level; and {A)Echinometra {E. lucunter + E. viridis) counts were added to the belt transects in all but the deep Isla de Sacrificios site. As the surveys were performed during the local rainy season, water turbidity was high. Visibility at any depth was usually <5 m, and often <2 m. In order to standardize the data, we spent a day assessing the same transects and quadrats until all team members consistently recorded similar results. RESULTS Stony Corals A team of seven divers (four divers/dive) surveyed 723 large (>25 cm in diameter), living stony corals, examined 662 quadrats, and conducted 138 live stony coral cover transects plus 124 belt transects for sea urchin abundance. Live stony coral coverage varied from 15% in Galleguilla to 21% in Isla de Sacrificios, averaging 17% overall (Table 1). The remaining substratum mostly consisted of coral skeletons that were moderately intact to highly eroded. We recorded 14 taxa of large living scleractinians and one hydrozoan (Table 2). The total number of taxa per reef varied from 6 in Galleguilla to 14 in Isla de Sacrificios, and from 12 in the deeper zones to 13 in the shallow zones. In terms of numerical abundance, Montastraea cavernosa was the predominant large species (overall mean = 35%), dominating both depths in Isla de Sacrificios and the shallow Galleguilla transects. Colpophyllia natans was the second most common overall (26.5%) and dominated both depths in Isla Verde. Siderastrea was the most common large taxon in the deep Galleguilla site. Depth had relatively little effect on mean numerical abundance with M cavernosa > C. natans > Diploria > Siderastrea in the three shallow zones, and M cavernosa > C. natans > Siderastrea > Diploria in the three deeper zones. The average numbers of large living stony corals per 10 m transect varied from 5 in Galleguilla and Isla Verde to 6.5 in Isla de Sacrificios (Table 1). Their mean size (as maximum diameter) overall was 59 cm (Table 3). Bleached or partially bleached corals accounted for 3% of the overall total. Black-band disease (in Siderastrea, Montastraea spp.), tumor neoplasms (in Diploria) and dark spots disease (in Siderastrea) were seen in the VRS during the survey but no diseases were found in any of the surveyed corals. Recent partial-colony mortality (hereafter recent mortality) of the large live stony corals averaged <1% of their upward-facing surfaces in five sites and was greatest (mean = 3%) in the shallow Isla de Sacrificios site. The percentage of affected corals varied from 0 to 13%, averaging 4% overall (Table 3). Average values for old partial-colony mortality (hereafter old mortality) varied between 6.5% at Galleguilla shallow and 13% at Isla Verde 364 shallow, with over 40% of all colonies having areas of old tissue loss. Trends in total (recent + old) mortality were similar to those for old mortality. Coral recruitment averaged 0.1/0.0625 m^ (~1.2/m^) and was low in all sites (Table 4). Sixty-seven recruits were counted: the numbers/taxa (and percent of total) were Siderastrea 43 (64%), Oculina 12 (18%), Porites and Agaricia 3 each (4%), M annularis and C. natans 2 each (3%), S. intersepta and M complanata 1 (1%) each. Algae and Echinoids In general, crustose coralline algae (overall absolute abundance = 41%), followed by turf algae (overall absolute abundance = 26.5%), dominated the algal quadrats; however, turf algae were somewhat more abundant than crustose corallines at both depths off Galleguilla (Table 4). Macroalgae were notably scanty in the VRS fore-reef zones (overall absolute abundance = 0.5%) and, as their mean height was <1 cm, the absolute macroalgal index (macroalgal abundance x macroalgal height) was approximated as 0 to <1. On average, “barren areas” covered with sediment (terrigenous + endogenous) and lacking conspicuous sessile organisms occupied 13.5% of the substrata and, at the Isla de Sacrificios deeper site, they dominated the available substratum (60%). Although present in the VRS, only two specimens of Diadema antillarum were seen during the survey and none were counted in the belt transects. Densities of Echinometra varied between lO/lOOm^ in the deep Galleguilla site and 155/100 m^ in the shallow Isla de Sacrificios site, averaging 65/lOOm^ overall (Table 4). DISCUSSION Live stony coral cover was much higher during the mid 1960's when Kiihlmann (1975) recorded values of 50% in shallow and 40% in deep areas, respectively, of Blanquilla reef At that time, Acropora. palmata covered up to 65% of the available substrata in some shallow VRS reefs and the cover of A. cervicornis reached 100% in Enmedio reef (Ranefeld, 1972; Kiihlmann, 1975). Live Acropora spp. accounted for <1.5% of the stony coral cover at 3-6 m in 1999. Overall, our data indicate that the VRS reefs had changed little since TunnelTs (1992) report on their condition when coral cover was 1 7% and 1 2% in the fore reefs at Enmedio and Cabezo Reefs, respectively. The reduced coral cover in the VRS reefs is a clear indication of serious ecosystem decline, even when it is recognized that nowadays an average of 1 7% is fairly typical of many Caribbean reef areas (see Kramer, this volume). That the VRS mortality estimates for colony surfaces in large (>25 cm diameter) stony corals are low (especially for old mortality) in comparison to data collected elsewhere in the western Atlantic (Kramer, this volume) is partially explicable by our exclusion of standing dead corals from these assessments. Nevertheless, in spite of the scarcity of coral diseases and the relatively high availability of crustose coralline algae as a substratum for larval settlement (Morse et al., 1994), the chances for significant recovery of the VRS fore-reef communities seem poor so long as coral recruitment, especially by the broadcasting species that are so important in reef framework construction, is low. 365 ACKNOWLEDGMENTS The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. Special thanks are extended to the Veracruz- AGRRA Survey Team: the field work would not have been possible without the assistance of Guadalupe Barba, Gilberto Acosta, Xochitl Gonzalez, Kublai Horta, Eduardo Palacios and Edgar Tovar. A very special thanks to Judy Lang for help with the revision of this manuscript. REFERENCES Beltran-Torres, A.U., and J.P. Carricart-Ganivet 1999. Lista revisada y clave para los corales petreos zooxantelados (Hydrozoa: Milleporina; Anthozoa: Scleractinia) del Atlantico mexicano. Revista de Biologia Tropical 47(4):8 13-829. Carricart-Ganivet, J. P., and G. Horta-Puga 1993. Arrecifes de coral en Mexico. Pp. 81-92. In: S.I. Salazar-Vallejo, and N.E. Gonzalez (eds.). Biodiversidad Marina y Costera de Mexico. Comision Nacional para el Conocimiento y Aprovechamiento de la Biodiversidad and Centro de Investigaciones de Quintana Roo. Mexico. Diario Oficial 1 992. Decreto por el que se declara area natural protegida con el cardcter de Barque Marino Nacional, la zona conocida como Sistema Arrecifal Veracruzano, ubicada frente a las Costas de los municipios de Veracruz, Boca del Rio y Alvarado del estado de Veracruz Llave, con superficie de 52,238-91-50 hectdreas: Diario Oficial de la Federacion, Mexico (24 de Agosto de 1992): 6-15. Gutierrez-Carbonell, D., C. Garcia-Saez, M. Lara, and C. Padilla 1993. Comparacion de arrecifes coralinos Veracruz y Quintana Roo, Pp. 787-806. In: S.I. Salazar-Vallejo, and N.E. Gonzalez (eds.). Biodiversidad Marina y Costera de Mexico. Comision Nacional para el Conocimiento y Aprovechamiento de la Biodiversidad and Centro de Investigaciones de Quintana Roo. Mexico. Horta-Puga G, and J.P. Carricart-Ganivet 1993, Corales petreos recientes (Milleporina, Stylasterina y Scleractinia) de Mexico, Pp. 66-79. In: S.I. Salazar-Vallejo, and N.E. Gonzalez (eds.). Biodiversidad Marina y Costera de Mexico. Comision Nacional para el Conocimiento y aprovechamiento de la Biodiversidad and Centro de Investigaciones de Quintana Roo. Mexico. Jordan, E. 1992. Recolonization patterns of Acropora palmata in a marginal environment. Bulletin of Marine Science 5:104-1 17. Kuhlmann, D.H.H. 1975. Charakterisierung der korallenriffe vor Veracruz/Mexiko. Internationale Revue der gesamten Hydrobiologie 60:495-521. 366 Lang, J., P. Alcolado, J.P. Carricart-Ganivet, M. Chiappone, A. Curran, P. Dustan, G. Gaudian, F. Geraldes, S. Gittings, R. Smith, W. Tunnell, and J. Wiener 1998. Status of coral reefs in the northern areas of the wider Caribbean. Pp. 123- 134. In C. Wilkinson (ed.). Status of Coral Reefs of the World: 1998. Australian Institute of Marine Science. Australia. Lara, M., C. Padilla, C. Garcia, and J. J. Espejel 1992. Coral reefs of Veracruz Mexico 1. Zonation and Community. Proceedings of the Seventh International Coral Reef Symposium. Guam 1:535-544. Morelock, J., and K.J. Koenig 1 967. Terrigenous sedimentation in a shallow water coral reef environment. Journal of Sedimentary Petrology 37:1001-1005. Morse, D. E., A. N. C. Morse, P. T. Raimondi, and N. Hooker 1994. Morphogen-based chemical flypaper for Agaricia humilis coral larvae. Biological Bulletin 186:172-181. Ranefeld, J.W. 1972. The stony corals of Enmedio Reef off Veracruz, Mexico. M.S. Thesis. Oceanography. Texas A&M University. Colleges Station, Texas. 104 pp. Tunnell, J.W. 1992. Natural versus human impacts to Southern Gulf of Mexico coral reef resources. Proceedings of the Seventh International Coral Reef Symposium, Guam, 1 :300-306. Vargas-Hemandez, J. M., A. Hemandez-Gutierrez, and L.F. Carrera-Parra 1993. Sistema Arrecifal Veracruzano. Pp. 559-575. In: S. 1. Salazar-Vallejo, and N.E. Gonzalez (eds.). Biodiversidad Marina y Costera de Mexico. Comision Nacional para el Conocimiento y Aprovechamiento de la Biodiversidad and Centro de Investigaciones de Quintana Roo, Mexico. 367 Table 1. Site information for AGRRA stony coral and algal surveys off Veracruz, Mexico. Reef name Site Reef type' Latitude Longitude Survey Depth Benthic >25 cm live % livestony code (o-N) (0... w) date(s) (m) Transects stony corals coral cover (#0 (#/10mf (mean ± sdf Shallow Galleguilla GA3 Wind fore 19 13 53 96 07 37 July 20-22 99 3-6 24 4.5 16.0 ± 11.5 Isla de Sacrificios IS3 Wind fore 19 1026 96 05 32 July 18-1999 3-6 22 7.0 19.5 ±22.5 Isla Verde IV3 Wind fore 19 11 50 9604 06 July 15 99 3-6 28 4.0 14.5 ±9.0 All shallow VR3 3-6 74 5.0 16.5 ± 15.0 Deeper Galleguilla GA9 Wind fore 19 13 53 96 07 37 July 19-20 99 9-12 24 5.0 14.0 ±8.0 Isla de Sacrificios IS9 Wind fore 19 1026 96 05 32 July 16-17 99 9-12 19 5. 5 22.5 ± 15.0 Isla Verde IV9 Wind fore 19 11 50 96 04 06 July 14-1699 9-12 21 6.0 18.5 ±8.0 All deeper VR9 9-12 64 5.5 18.0 ± 11.0 All Galleguilla AIIGA 3-6; 9-12 48 5.0 15.0 ± 10.0 All Isla de All IS 3-6; 9-12 41 6.5 21.0 ± 19.5 Sacrificios All Isla Verde All IV 3-6; 9-12 49 5.0 16.0 ± 8.0 All sites All VR 3-6; 9-12 138 5.0 17.0±13. 5 'wind = windward ^Standing dead corals were not surveyed. Table 2. Relative abundance and species richness of all live* stony corals (>25 cm diameter) by site off Veracruz, Mexico. Species Galleguilla % Isla de Sacrificios % Isla Verde % All Reefs % 3-6 m 9-12 m 3-6 m 9-12 m 3-6 m 9-12 m 3-6m 9-12 m Both depths M alcicornis 1.0 3.5 1.0 1.0 1.5 0.5 1.0 A. cervicornis 1.0 0.5 <0.5 A. palmata 1.5 9.0 3.0 1.5 Agaricia spp. 0.5 2.0 0.5 0.5 0.5 C. natans 20.5 10.0 12.5 9.0 50.0 60.0 25.5 27.5 26.5 Diploria 4.5 1.0 19.5 5.0 23.0 180 16.0 8.5 12.5 M. decactis 0.5 0.5 <0.5 M. annularis 6.0 5.5 4. 5 1.5 3.5 2.5 M. cavernosa 59.5 30.0 49.0 56.0 7.5 8. 5 40.0 29 5 35.0 M. faveolata 6.0 2.0 1.0 0.5 2/0 1.0 M. franksi 1.0 0.5 <0.5 0. diffusa 0.5 0.5 <0.5 P. astreoides 2.0 1.0 1.0 0.5 0.5 Siderastrea 13.5 58.5 10.5 11.0 3.0 4. 5 9.0 25.5 17.0 S. intersepta 2.0 4.0 1.0 0.5 1.5 1.0 All taxa (#) 5 5 10 9 8 10 13 12 15 'Standing dead corals were not surveyed. 368 o o ’>< '(U o cd o (U 0) .2 '"B E o »n (N Al C3 O o c o 03 o _o ’■*— > 2 ’> q/D +1 C o3 (U c _o '■*— • c o o "O c 03 ^ CQ H o 2 o i E c •:= o t- E V B S 5 o o o o O (N rf 2 rr; in ^ -H in 2 Qi m p in in o V rd o ro 4^ 41 4^ m On NO in in m On — in o 4^ ^ < C? O m > fn > o o o o in o in ^ p •* 41 in o o p fn p p p On rd NO ^ < p in Tf ^ m 00 fs in in in in o rn 00 41 41 o m p p in ri in 41 41 41 p in p 00 On Os 41 in o ^ o fn m m 41 41 41 ON in ^ m m NO Tf NO fn m m fS fS fS i = = < < < TD > £ 3 O c fli c/i uL (D V “O ? o c/5 T3 (2 « 1) •O 0) ex) X) C C3 ^ H § 5^ 369 o p3 X) (U t > _c o 0) t; d) o cc3 cc3 X O — t/3 '^" .2 JJ 'eS -2 ’> d) H "O 370 Figure 1. AGRRA survey sites (Kalki, Westpunt, Jeremi, Lagun, Oostpunt) on the leeward coast of Cura9ao. Percentages = mean percentage of diseased stony corals (all species and all sizes) in belt transects at 10 m, 15m and 20 m that had been found in 1 997 at these sites, and at four other reefs. CONDITION OF CORAL REEFS OFF LESS DEVELOPED COASTLINES OF CURACAO (PART 1: STONY CORALS AND ALGAE) BY ANDREW W. BRUCKNER* AND ROBIN J. BRUCKNER^ ABSTRACT Coral reefs at 10-20 m depth off eastern and western Cura9ao, Netherlands Antilles had high abundance and high cover (25-50%) of stony corals, although the latter declined between 1998 and 2000, primarily from impacts associated with Hurricane Lenny in November 1999 and coral disease. Most corals had lost 15- 40% of their live tissues and the amount of partial mortality declined with depth. Little recent mortality was observed (0.6% in 2000). Reefs were dominated by the Montastraea annularis species complex (46% of all corals >20 cm in diameter), which were 40% larger than other species. Overall, colonies of the M. annularis species complex sustained somewhat greater total (recent + old) partial-colony mortality (24%) than other stony corals (19%), and had a higher prevalence of disease. Yellow-blotch disease affected 14.5% of all colonies of the M. annularis species complex in January 2000; infected corals had twice as much total partial- colony mortality (44%) as uninfected conspecifics. Shallow reef communities at 8- 12 m appear resilient to disturbance, as evidenced by low macroalgal cover, a high abundance of stony coral recruits and juveniles, and declining disease incidence and prevalence overall. However, the high incidence of yellow-blotch disease in the M. annularis species complex and the absence of recruits of these species suggests their condition may continue to decline and a shift in species dominance may be underway. INTRODUCTION Cura9ao, loeated 60 km north of Venezuela, forms part of the leeward Netherlands Antilles. The small oceanic island (61 km long; 443 km total) is surrounded by fringing coral reefs, which are better developed along the leeward coast. The shallow reef community, which begins 20 m to 250 m from the shoreline, consists of a terrace that slopes gradually seaward to 7-13 m depth and ' Corresponding author, NOAA/National Marine Fisheries Service, Office of Protected Resources, 1315 East West Highway, Silver Spring, MD 20910, USA. Email: Andy.Bruckner@noaa.gov 2 NOAA/National Marine Fisheries Service, Office of Habitat Conservation, 1315 East West Highway, Silver Spring, MD 20910 USA. 372 then slopes steeply to a sand terrace at 50-60 m (Pors and Nagelkerken, 1998). Twenty-five years ago, the vertical reef profile was characterized as having a shallow palmata zone dominated by elkhom coral and gorgonians, fields of Acropora cervicornis from 4-5 m depth, and a reef slope dominated by the Montastraea annularis species complex, Agaricia spp. and Madracis mirabilis (Bak, 1975). Coral cover and diversity were high on the reef slope, but decreased rapidly below 35-40 m. Fifty-seven species of scleractinian corals were identified by Bak (1975). The coral reefs surrounding Cura9ao are affected by a number of natural and anthropogenic stressors. The island is located south of the hurricane belt and rough seas are rare on the leeward coast. However, tropical storms pass within 200 km of the island about every four to five years, and associated wave surge has caused considerable damage to the shallow reefs (Van Duyl, 1985; Van Veghel and Hoetjes, 1995). Development and industry are concentrated in Willemstad and along the adjacent southeastern, leeward coast where the majority of Cura9ao’s population (155,000) resides. In these developed areas coral abundance, cover, and species diversity declined precipitously at 1 0-20 m on fore reefs between 1973 and 1992 (Bak and Nieuwland, 1995). Much of this change was attributed to sewage discharge and to sedimentation associated with beach construction (Bak and Nieuwland, 1995). The island-wide mass mortality of the herbivorous sea urchin Diadema antillarum in 1983 also contributed to a general decrease in the cover of live corals and coralline algae, while turf algae and macroalgae increased in abundance (Bak et ah, 1984; De Ruyter van Steveninck and Bak, 1986). Branching acroporids (A. palmata, A. cervicornis) suffered high mortality in 1980 and 1981 from white-band disease (WBD) (Bak and Criens, 1981; Van Duyl, 1985), but other stony coral diseases were only minor sources of mortality (Bak and Nieuwland, 1995). Bleaching events occurred in 1987 (Williams and Bunkley-Williams, 1990), 1990 (Meesters and Bak, 1993), 1995 (CARICOMP, 1997) and 1998 (A. Bruckner, unpublished data). Yellow-blotch disease (YBD) was first noticed in late 1995 as colonies of the M annularis species complex began to recover from the mass bleaching event. It is not known whether colonies with YBD had bleached during this event (P. Hoetjes, pers. comm.). While Cura9ao’s reefs clearly have degraded near its population center, its eastern and western coasts are relatively unaffected by pollution or sedimentation and their reefs are thought to be in better condition (Van Veghel, 1997). Subsistence fishers occur throughout the island, and although spearfishing and coral collection were prohibited in 1976 regulations were not enforced until 1998. The Cura9ao Underwater Park, established in 1983 and extending 21 km from the outskirts of Willemstad to the eastern tip of the island, includes a 12 km stretch of coastline that is undeveloped and uninhabited. In total, the park encompasses 600 hectares of coral reef habitat and 436 hectares of inner bays. Hook-and-line fishing is allowed within the park, but spearfishing and anchoring are prohibited. All dive sites have mooring buoys. A Caribbean Coastal Marine Productivity CARICOMP) coral reef site has been monitored here at Spaanse Water since 1994. Lagun and Westpunt are two small communities at the western end of Cura9ao where there is 373 no industry and very limited coastal development. A second underwater park has been proposed for these reefs (Banda Abao reef complex). The purpose of this study was to utilize the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol to characterize the reefs off the eastern and western ends of Cura9ao, in particular to determine whether diseases are an important factor contributing to coral mortality in the absence of significant human activities. Our data suggest that coral diseases have become more prevalent on Cura9ao’s reefs during the late 1990s and that one disease in particular is causing significant partial mortality to the most abundant and most important of its reef- building corals. Another unexpected source of mortality was attributed to wave energy associated with Hurricane Lenny in October, 1999. Fish assessments made in June 2000 are reported separately in Bruckner and Bruckner (this volume). METHODS Baseline data were initially obtained in June 1997 by tallying the total number of healthy and diseased colonies of each species of stony coral (scleractinian corals and hydrozoan fire corals) observed within belt transects (2 m wide X 30 m long; 1-4 transects/site) along depth gradients (at 10, 15, 20 m) for nine leeward reefs (Fig. 1). In August 1998 and January 2000, detailed reexaminations were conducted at strategically chosen sites on the less populated eastern and western coasts using the AGGRA Version 2 benthos protocol (Appendix One, this volume), with the following modifications. The minimum diameter of assessed corals was 20 cm and size measurements were recorded to the nearest 5 cm. Smaller colonies of reef-building corals (5-20 cm) were tallied and recorded to species. Scleractinian recruits, defined as <2.5 cm diameter, were recorded to genus only (except for Montastraea cavernosa), omitting taxa that do not obtain a large size (e.g., Scolymia spp., Favia fragum). Colonies of the Montastraea annularis species complex were separated according to Weil and Knowlton (1994) as M. annularis, M. faveolata or M. franksi. Forms or morphotypes of Agaricia agaricites, Colpophyllia natans, Meandrina meandrites and Porites porites were combined under the respective species. Encrusting forms of Madracis were recorded as M. decactis. Recent mortality was defined in this study as any tissue loss occurring within approximately the last 60 days, and signs included: (1) white coral skeleton that lacked algae (surfaces denuded of tissue within the last five-seven days); (2) skeletal areas with readily recognizable corallites that had not been substantially eroded but were colonized by green filamentous algae; or (3) white, exposed skeletal surfaces, or eroded skeletal surfaces with fine filamentous algae that had been physically abraded by fish or other agents but had not yet been colonized by macroalgae or coralline algae. Causes of recent mortality were identified as disease [separated into WBD, YBD, black-band disease (BBD), white plague (WP), dark-spots disease (DSD) or other diseases], corallivory [fish bites, damselfish (Stegastes planifrons) algal lawns, or snail predation], overgrowth by algae or an invertebrate (cnidarian, sponge or tunicate), or were recorded as unknown. In January 2000, when toppled corals were observed throughout the reefs, especially at 8-12 374 m, colonies that had become stabilized were measured and described as above, but were tallied as overturned. The long-dead portions that had been recently exposed through toppling were recorded as old mortality; only skeletal areas that met the criteria described above as recent mortality were tallied as sueh. Dislodged or overturned corals that had not become stabilized were not included in this survey. To standardize observations, all measurements of mortality were performed by the first author and algal quadrats were completed by the second author. Both authors collected information on colony size. All corals along two pilot transects (Lagun and Jeremi, 10 m length) were measured by both authors and measurements were discussed and compared to ensure consistency prior to the actual surveys. All statistical analyses were performed with the Systat (version 9.0) program. Comparisons among species, locations, and depths were made using a student’s Mest (for examination of the M annularis species complex versus all other species pooled) or ANOVA (single-factor or two-factor) and correlations were examined with a Pearson product-moment test. A one- or two-factor ANOVA was also used to examine for differences in coral composition, size frequency distribution and percent partial mortality, surveys from different years on western reefs or between eastern and western reefs. When relevant, post-hoc analyses were performed using a Tukey HSD multiple comparison test. For these tests, coral species were lumped into the following six groups based on colony abundance, mean colony size, susceptibility to disease or predation, colony morphology or sexual reproductive character: (1) C. natans; (2) the M annularis species complex; (3) other broadcast spawners with massive morphologies {Diploria spp., M cavernosa, Siderastrea siderea, Stephanocoenia intersepta); (4) small branching corals {P. porites species complex, Eusmilia fastigiata and Madracis spp.); (5) Agaricia spp.; and 6) other species {Porites astreoides, M. meandrites, Mycetophyllia spp., Mussa angulosa). Data were checked graphically to assure that all assumptions of ANOVA were met; log-transformation for length measurements and arc-sine transformation for percentages were used as appropriate prior to analyses. RESULTS Stony Corals AGRRA surveys were performed at 9- 1 3 m on the reef terrace at one eastern (Oostpunt) and three western (Kalki, Jeremi, Lagun) sites during August 1998 (Table 1). Further surveys were made at four western sites (Kalki, Westpunt, Jeremi, Lagun) two months after Hurricane Lenny in January 2000, at ~10 m on the reef terrace and to a maximum depth of ~20 m on the reef slope. Reefs at 10-20 m were characterized by 18-24 species of scleractinian corals (at least 20 cm in diameter) and the hydrozoan Millepora complenata, but were dominated by the Montastraea annularis species complex, Agaricia agaricites, Montastraea cavernosa, Colpophyllia natans (Fig. 2) and, at 15-20 m, by A. lamarki and Stephanocoenia intersepta. In 1998, species composition at 10 375 m in the six pooled coral groups (see Methods) did not differ among locations (ANOVA, p=0.09). Numerically the most abundant corals at all sites and depths (46% of total) belonged to the M annularis species complex which collectively are the primary live cover and structural element of Cura9ao’s fringing reefs. At the western sites 70% of all corals at 10 m depth consisted of the M. annularis species complex, with M faveolata >M. annularis>M. franksi; 35-42% of all corals at 15 and 20 m (respectively) consisted of these species, with M faveolata>M. franksi > M. annularis. Similarly, 40% of the corals examined at 10 m in Oostpunt were M. annularis and M. faveolata. Small, isolated colonies of Acropora cervicornis (staghorn coral) and numerous patches of dead staghorn rubble were identified between colonies of the M. annularis species complex within transect areas at Oostpunt. On western reefs, the substratum at 7-10 m depth often consisted of dead, consolidated staghorn rubble, but live colonies of A. cervicornis were not observed within transect areas nor in the surrounding reefs. Shallow areas (2-4 m) outside of transect areas at Oostpunt had a low abundance of live A. palmata. This species was rare or absent in other locations, and the shallows (0-5 m) were nearly devoid of living coral. Large patches of Madracis mirabilis, a number of which were several meters in diameter, occurred in 10-15 m on the Kalki reef, and less frequently at other locations. Colony density (for stony corals of >20 cm diameter) generally ranged from 1. 3-2.1 corals per meter (Table 1); data from 2000 indicates that colony density increased with depth (r^=0.61; p=0.0016). Coral cover varied among locations, depths and years. In 1998, coral cover ranged from about 20-40% along transects at 10 m depth and was greatest off Kalki. Coral cover at the same depth was substantially lower on the western reefs in January 2000, except for Jeremi, where high variation among the transects, and the presence of several large colonies of M faveolata, may have skewed its mean value (Table 1). The greatest decline in live cover was observed at Kalki where most corals had been removed at 2-12 m from the reef terrace by wave surge. Overall, coral coverage was greater at 15-20 m on reef slopes (Table 1) with the highest percentage occurring at 20 m on Jeremi Reef (49%). Live coral cover was lowest off Westpunt even though this site was minimally impacted by Lenny. Westpunt Reef terminates in sand at 15 m and living corals are absent below this depth. Transects performed on western reefs in 1998 and 2000 were similar in composition (two-factor ANOVA; p=0.59) and diameter of corals (p=0.54), with no interaction between years and species (p=0.79) with respect to coral size. Colony size recorded during 1998 in the six pooled coral groups did not differ among locations (ANOVA, p=0.91). Coral diameter did, however, differ among 376 A M. cavernosa Agaricia spp. M. franksi M. annularis 10 m (n=358) S. siderea Porites spp. S. intersepta C. natans other M. faveolata B M. cavernosa M. franksi Agaricia spp. 15 m (n-616) other S. siderea Porites spp. S. intersepta M. faveolata C. natans M. annularis Figure 2. Species composition and mean relative abundance of all stony corals (>20 cm diameter) in January 2000 at (A) 10 m, (B) 15 m, (C) 20 m, in western Curafao. 377 A Size class (cm) B Size class (cm) Figure 3. Size-frequency distribution of colonies (>20 cm diameter) in Janaury 2000 of (A) Montastraea annularis (10 m, 15 m) and M franksi (at 15 m, 20 m), (B) M faveolata (at 10 m, 15 m, 20 m) in western Curasao. 378 the species groups (two-factor ANOVA; P<0.001) and an interaction between diameter and location was observed (P="0.009). The largest colonies observed within transect areas were M. faveolata and C. natans, while most of the brooding species (e.g., Agaricia spp. and Porites spp.) showed a predominance of the smaller size classes. In January 2000, a large number of the smaller colonies, especially M. annularis, were overturned or displaced. The effects of the storm were highly localized, however, as deeper areas and larger corals were minimally impacted. Most corals examined in January 2000 were intermediate in size (30-80 cm diameter, mean=53 cm, n=1501; Fig. 3). Colonies of the M. annularis species complex were significantly larger in diameter (mean=63 cm) than all other species combined (mean=45 cm) (all depths pooled; t-test, t=12.5, df=1497, p<0.001). In addition, colonies of M. faveolata and M. franksi (mean=67 cm) were larger than M annularis (mean=46 cm; ANOVA, MS==1.13, F=24.7, p<0.001), but no differences were noted among M. faveolata and M franksi. Colony size did not differ among depths for the M. annularis species complex (ANOVA, MS=0.1 1, F=2.35, p=0.095) or other species (ANOVA, MS=0.03, F=1.01, p=0.36). Small corals (5-20 cm diameter) recorded in 1998 along transects (mean abundance=0. 5/meter, range=3-9/10 m) consisted predominantly of A. agaricites (1.3/10 m), P. porites (0.6/10 m), M annularis (0.5/10 m), M meandrites (0.5/10 m), M mirabilis (0.5/10 m) and 16 other species (Fig. 4A). A low abundance of recruits (<2.5 cm diameter) was identified within quadrats at 10 m depth (0.5 - 0.7 recruits/0.0625 m ) in 1998 (Table 3), most individuals of which were brooders including Agaricia and Porites. However, broadcasters like M cavernosa, Dichocoenia, Colpophyllia, Stephanocoenia and Meandrina were also observed (Fig. 4B). An absence of recruits of the M. annularis species complex was noted, even though these were the dominant corals on all reefs. Recruits were also recorded in 2000 but lower numbers were seen (mean=0.12 recruits/0. 0625m^ at 10 m depth). In shallow transects (9-12 m), recruits were not observed on reef substrata with a high cover of macroalgae (Jeremi and Lagun) or on substrates that had been exposed relatively recently (e.g., at Kalki). Recruits were observed on long-dead coral skeletons and reef substrata not directly affected by the hurricane with a higher number at 15 m on reef substrata (up to five recruits/0. 0625m^ mean=0.61) than on denuded coral (especially M. faveolata, M. annularis, M. franksi) skeletons (mean=0.13) (t test, /=3.5; df=108; p<0.001). Coral Condition In this study we examined the condition of 1,939 scleractinian and hydrozoan reef-building corals (1998 and 2000, all sites and depths pooled). Overall, in 32% of all corals total (recent + old) partial-colony mortality (hereafter total partial mortality) affected less than 1 0% of their planar surface area. Mean values of total partial mortality at each site (all corals pooled, 1998 and 2000) ranged from 15-32% with less than 20% of all corals missing more than half their tissues (Fig. 5). Distinct differences in percent total partial mortality were also noted among the six species groups (all years and depths are pooled, p<0.001). 379 A P. astreoides Juveniles (n=150) Millepora other P. porites Agaricia C. natans D. stokessi Diploria Eusmilia Meandrina M. cavernosa M. annularis B Recruits (n=71) Porites Stephanocoenia Agaricia Colpophyllia Meandrina Dichocoenia M. cavernosa Figure 4. Species composition and mean relative abundance in August 1998 of (A) all “juvenile” (5-20 cm diameter) stony corals and (B) all recruits (<2.5 cm diameter, excluding species that are small as adults) at 10 m, 15 m and 20 m in western (four sites) and eastern (one site) Curasao. 380 t; 03 CX o + c c D o u. 0) CL M. annularis species complex 10 m (n=251) ■ 15m(n=259) □ 20m(n=187) <10 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-100 Percent partial mortality (old and recent) S 60- o E >. J 50- o o 40 •• other species ii 10m(n=lll) I 15 m (n=355) □ 20 m (n=338) <10 10-19 20-29 30-39 40-49 50-59 60-69 70-79 80-89 90-100 Figure 5. Frequency distribution in January 2000 of total (recent + old) partial colony mortality of all colonies (>20 cm diameter) of (A) the Montastraea annularis species complex, (B) other species, in western Cura9ao. 381 From 20-40% of each small clump of^. cervicornis consisted of dead branches. Colonies of the widely distributed M annularis species complex overall exhibited a significantly greater amount of total partial mortality than all other species (pooled) except A. cervicornis (all sites, depths and years are pooled, mean partial mortality=24%; t-test, t=62, df=1497, p<0.0001). No relationship was observed between percent tissue loss and colony diameter for the M annularis species complex (r = 0.002; p=0.28) or other pooled species (r =0.04), possibly due to the high amount of variation observed within each size class. In 1998, the amount of total partial mortality in the six pooled species groups was found to vary among species (two-factor ANOVA; P<0.001), with minor differences among locations (p=0.066) and a significant species-location interaction with respect to total partial mortality (p<0.001). Total partial mortality was greatest among the M. annularis species complex and branching corals (Porites porites, Eusmilia fastigiata, Madracis spp.) while the group of brooding species had the lowest percentage (Tukey test). Post-hoc analysis indicates that total partial mortality in 1998 was slightly higher on western reefs (21% versus 19% at Oostpunt), yet the greatest amount overall (35%) was recorded for the M annularis species complex at Oostpunt. In 2000, a large number of small-to-intermediate-sized corals were dislodged or overturned in shallow water (7-13 m depth), and numerous unattached colonies had been transported down the reef slope to 15-20 m or deeper. Overturned colonies identified along transects that were 20 cm or larger had a mean size of 49.5 cm (maximum=160 cm). Overturned colonies were substantially larger on Jeremi (mean=60.5 cm) than on Lagun (mean=39.5 cm), but the total proportion of overturned colonies versus those that were unaffected was higher at Lagun (26%) than on Jeremi (20%). A small number (<3% of all corals examined) experienced total partial mortality; most of the corals that survived had sustained a low amount of recent partial-colony mortality (hereafter recent mortality) with the exception of areas on the colony now in contact with the substrata. The amount of total partial mortality in 2000 varied significantly among depths for the M. annularis species complex (ANOVA, MS=5628, F=24.8, p<0.001) and other species (ANOVA, MS=4103, F=13.2, p<0.001) with the greatest loss at 10 m depth and the least at 20 m (Table 2). Whereas the amount of recent mortality at 10 m on western reefs was greater in 1998 than 2000 (3.7% versus 0.6% of colony surfaces, respectively), the amount of total partial mortality at 10 m was significantly greater in 2000 (mean=29%) than in 1998 (mean=21%; two-factor ANOVA, p=0.003). Minor differences in percent total partial mortality were noted among the six species groups (p=0.06) but there was no interaction between the survey period and species (p=0.175). Overall, S. intersepta, the M annularis species complex, S. siderea and E. fastigiata respectively exhibited the greatest amount of total partial mortality (22-31%), while the lowest values were observed in M. meandrites, D. strigosa, D. stokesii, P. astreoides and P. porites (10- 13%). The other dominant species (C. natans, M. cavernosa, and Agaricia spp.) exhibited 17-19% total partial mortality. The most common sources of recent mortality were coral diseases, in particular YBD, DSD, WP, WBD and red-band disease (Table 4). On all reefs, a high percentage of the M annularis species complex exhibited signs of YBD; this condition was not recorded in other taxa. WBD was only observed at the Oostpunt, 382 susceptible acroporid corals being absent at the western sites. A high prevalence of DSD was noted in 2000. It is not known whether DSD was common in 1998 because the affected species (S. intersepta, S. siderea) were predominantly recorded at 15 and 20 m, at depths which had not been surveyed in 1998. Coral diseases appeared to be most common in 1997 (Bruckner and Bruckner, 1999; unpublished data), and have since declined. The highest prevalence of YBD was observed along the east coast, near Willemstad (30.6% of the 264 colonies of the M. annularis species complex examined at Piscadera; 49%, N=165 at Jan Thiel) and at Oostpunt (37.5%, N=474), with fewer infections seen off the western coast (24%, N=607). In 1998 fewer infections were observed in western sites at 10 m depth (15.6-19.2% of the M. annularis species complex) while Oostpunt had a higher prevalence of YBD (68% of all colonies of the M. annularis species complex). Although the occurrence of new YBD infections was lower in January 2000 on western reefs, many older infections still plagued colonies (total=14.5% of 698 of the M annularis species complex) as indicated by the large amount of partial mortality immediately adjacent to the YBD-affected tissues. Moreover, colonies of the M annularis species complex with YBD had lost a significantly greater percentage of their tissues than had unaffected conspecifics (mean loss=44% versus 20%, respectively; t-test, /=10.6, df=127, p<0.0001; Fig. 6A) and were larger in size (mean diametei^72 cm) than uninfected corals (61 cm). Diameter was correlated to incidence of YBD (r^=0.45, p=0.0001): 25% of all colonies greater than 0.5 m diameter were infected, versus only 7% of the smaller colonies (Fig. 6B). BBD and WP were also more prevalent during the summer of 1998 (3.6%) than in January 2000 (0.1%); however, these differences may relate to seasonal variations associated with water temperature rather than an overall decline in disease. An additional source of coral mortality was attributed to predation. Stoplight parrotfish {Sparisoma viride) bites were observed on all reefs. Focused biting was primarily observed among M. annularis and C. natans and spot-biting affected these and 12 other speeies. Overall, the most extensive skeletal and tissue destruction from S. viride occurred on M. annularis (9.8% of all colonies) at 10 m in Oostpunt in 1998. Lesions from fish bites were very common in 1997 and 1998 (4.4% of all colonies of M. annularis), and affected 2.2% of all colonies examined at 10 m depth in 1998. Although fewer colonies appeared to be affected by fish predation on western reefs in 2000, affected colonies were highly aggregated and areas exhibiting focused biting in 1998 also had affected colonies in 2000. Predation by the snail Coralliophila abbreviata (in 5/14 surveys) and overgrowth by sponges and the tunicate Trididemnum solidum (4/14 surveys) were also noted (Table 4). Bleaching was not recorded during the August 1998 and January 2000 surveys. However, bleached colonies were observed in November 1998, primarily among Agaricia spp. and the M annularis species complex (personal observations). During this bleaching event about 30% of the M. annularis species complex became pale, especially on their upper surfaces, but did not turn 383 Partial colony mortality (%) Figure 6. Relationship between the presence (or absence) in January 2000 of YBD and (A) percent of total (recent + old) partial colony mortality, (B) size frequency distribution, for the M. annularis species complex (>20 cm diameter) in western Cura9ao. Data are pooled from 10 m, 15 m and 20 m. Grey diamonds in (B) indicate the percent of each size class affected by YBD; a best fit line and r-value are presented. 384 completely white. Approximately 80% of all colonies of^. lamarki in deeper water (18-25 m depth) were completely bleached (white). Bleaching affected other speeies but was less prevalent. We tagged and photographed 30 of these corals {A. lamarki and the M. annularis species complex) in November 1 999 and reexamined them in January 2000. Over 75% (n=26) of these regained full pigmentation and did not experience any mortality. In the other four colonies, tissue mortality affecting 5-30% of the surface of each plate was noted but the remaining live tissues had regained most pigmentation. Algae Algal communities at 10 m in 1998 were dominated by crustose coralline algae (50 %) and sparse turf algae (38 %) that did not trap considerable amounts of sediment. Macroalgae were codominant on exposed coral skeletons and at the bases of coral heads (Table 3). Where present, macroalgae eonsisted primarily of Dictyota spp., Halimeda spp. and Lobophora variegata, and were usually less than 1.5 cm in height. The algal community was very different in January 2000, possibly due to the effects of Hurricane Lenny two months previously. At Kalki, the newly exposed reef substrata consisted primarily of cemented A. cervicornis skeletons that were being colonized by fine filamentous turf algae; macroalgae and erustose corallines were rare. Mats of cyanobacteria, which had been observed here and at Jeremi in August 1998, were uncommon at both sites. At Lagun, Jeremi and Westpunt, a dense growth of red algae (Liagora spp., Trichogloea spp., Trichogloeopsis pedicellata and other similar fleshy algae; mean height 5 ± 21.5 cm) occupied much of the exposed substrata in January 2000. On Jeremi, these algae occupied 43% of open substrata at 10 m depth but were not growing on living corals. Dead coral surfaces (i.e., areas on colonies denuded of tissue) had similar algal communities in 1998 and 2000 composed predominantly of fine filamentous turfs and crustose eorallines with sparse macroalgae. In most cases, patches of macroalgae {Dictyota spp, Lobophora spp. and Halimeda) occurred at the base of eoral heads, between lobes of living coral, and in crevices, while up to 80% of the exposed surfaces were colonized by turfs and crustose eoralline algae. Diadema antillarum was not observed along transects, although signs of grazing were apparent on exposed coral surfaces and herbivorous fish (mean densities in 1998: surgeonfish, 5/100 m^ and parrotfish, 4/100 m^) were reeorded in belt transeets (30 m long x 2 m wide) (Bruekner and Bruekner, this volume). DISCUSSION Coral reefs examined off Cura9ao in 1998 and 2000 were dominated by intermediate to large-sized massive scleractinian taxa, including M faveolata, M. annularis, M. franksi, S. intersepta, S. siderea and C. natans, and smaller colonies of P. astreoides and A. agaricites. Acropora palmata and A. cervicornis were prevalent in the 1970s, declined in the early 1980s as a result of a regional disease epizootic (Van Duyl, 1985), and were uncommon (Oostpunt, Jan Thiel, Piscadera) 385 or were not observed (western reefs) during these surveys. Living coral cover ranged from 17-49% with the exception of one shallow site on the western coast that lost >95% of its live stony corals as a result of a hurricane in 1999. Most colonies over 20 cm (69%) had experienced total partial mortality that affected at least 10% of their planar surface area, with a mean tissue loss in all colonies of 22%. The major sources of mortality noted in this study were coral diseases (prevalent in all sites and years) and hurricane damage (affecting western reefs in 2000). Diadema antillarum was rare or absent in all sites but a low abundance of macroalgae and turf algae suggests that herbivorous fish are effectively controlling algal populations. In general, the eastern reefs were in poorer condition than the western reefs. A precipitous decline in coral abundance, cover, and species diversity had been reported on reefs near Willemsted in previous studies (e.g., Bak and Nieuwland, 1995). In 1997, these areas had the highest prevalence of disease with up to 49% of the >20 cm diameter colonies affected in one location. Oostpunt, a protected area located off the uninhabited eastern end of Cura9ao, was reported to be in good shape as recently as 1995 (Van Veghl, 1997). This site experienced a mass bleaching event in the fall of 1 995 and present surveys revealed a high incidence of disease in 1997 (38%) and 1998 (21%) and other biotic disturbances (especially focused biting by S. viride). Colonies at this site had also sustained a higher percentage of total partial mortality, and there were more entirely dead colonies, than observed at the same depth on western reefs. The amount of total partial mortality differed among species, and was highest in the slowly reproducing, large, massive broadcasters (M. annularis species complex, S. siderea, S. intersepta, C. natans) that dominated eastern and western reefs. In particular, colonies of the M. annularis species complex, which were the most abundant and largest corals at all sites and depths, had experienced a high degree of total partial mortality. They were also affected most severely by disease: between 7-49% of all colonies were observed with signs of disease in surveys conducted in 1997, 1998 and 2000. The main disease affecting the M annularis species complex was YBD, a condition that causes relatively slow rates of mortality (1-2 cm spread per month), but may affect individual colonies for several years (Bruckner and Bruckner, 2000). The prevalence of large colonies of the M. annularis species complex that are hundreds of year old and few colonies less than 30 cm in diameter suggests that recruitment events of significance have not occurred among these species in several decades. These large, ecologically-dominant colonies have exhibited high rates of survivorship, and are presumably well adapted to deal with chronic disturbances such as predation, bioerosion and disease (Bythell et al., 1993). However, this may no longer be the case for these reefs as recent disease epizootics have primarily plagued these species. Overall, larger colonies of the M annularis species complex were infected with YBD more frequently than small colonies. In addition, colonies with active signs of YBD have lost a mean of 44% of their tissue area in January 2000 or roughly twice that of uninfected colonies of the same species. Chronic YBD infections on these reefs may have serious ramifications for the persistence of the M. annularis species complex, as continued 386 tissue loss and fission may significantly reduce the reproductive potential of these colonies and the proportion of small, non-breeding colonies in the population may increase. Furthermore, several decades or more may be required for their replacement because of their slow rate of growth, delayed reproduction, and infrequent, episodic recruitment (Szmant, 1991). Although coral reefs surveyed in this study have a high prevalence of coral disease and have been impacted by a recent hurricane, Curacao’s reefs appear to be relatively resilient to recent disturbances and have a high potential for continued growth and sexual recruitment. Crustose coralline algae are prevalent on exposed reef substrata; macroalgae and turf algae have remained sparse with the exception of a benthic algal bloom in shallow water that lasted for three months after Hurricane Lenny and has since disappeared. Most locations have relatively high cover of live coral (over 25%), a high abundance of small corals less than 20 cm (3-9/10 m) and new recruits are present (0.5-0. 7 recruits/.0625 m ). Longer-term effects associated with the hurricane appear to be minimal. Most colonies overturned during Lenny have become stabilized; fragmented and overturned corals contained substantial amounts of surviving tissue which were not diseased, bleached or injured. In addition, surveys from January 2000 indicate that the number of active and new YBD infections have declined and other diseases were less abundant. Reefs at the western end of the island have experienced coral mortality as a result of disease and hurricane damage but ensuing mortality appears to be declining, coral cover remains high and remaining corals are in relatively good condition. In light of an ongoing accelerated decline of reefs located near human population centers, including those off populated coastlines of southeastern Cura9ao (Bak and Nieuwland, 1995), conservation efforts should be directed towards the island’s more remote western end. The western reefs may serve as a refuge for important reef-building species and a source of larvae for these and other reefs down current. Recent prohibitions on spearfishing, along with the proposed establishment of a marine park for western Cura9ao, are two steps forward that may help ensure the long-term persistence of its reef ecosystems. ACKNOWLEDGMENTS This research was funded in part by a grant from the Professional Association of Dive Instructors (PADI) Project Aware. Reviews of earlier version of this manuscript by Tom Hourigan are gratefully acknowledged. This manuscript was greatly improved by extensive comments and suggestions provided by Judy Lang. We thank All West Divers for logistical support, including boat transportation and scuba tanks. The Carmabi Foundation (Carmabi: Caribbean Research and Management of Biodiversity) also provided logistical support for surveys conducted on the southeastern reefs. Reef Care Cura9ao provided boat transportation and scuba tanks for surveys at Oostpunt. Mr. Errol Coombs assisted with several aspects of the field research, and his help is greatly 387 appreciated. The opinions and views expressed in this document are those of the authors and do not necessarily reflect those of the National Marine Fisheries Service, National Oceanic and Atmospheric Administration or the U.S. government. REFERENCES Bak, R.P.M. 1975. Ecological aspects of the distribution of reef corals in the Netherlands Antilles. Bijdragen Dierkunde 45:181-190. Bak, R,P.M., M.J.E. Carpey, and E.D. De Ruyter Van Steveninck 1984. Densities of the sea urchin Diadema antillarum before and after mass mortalities on the coral reef of Cura9ao. Marine Ecology Progress Series 17:105-108. Bak, R.P.M., and S.R. Criens 1981. Survival after fragmentation of the colonies of Madracis mirabilis, Acropora palmata, andyl. cervicornis (Scleractinia) and the subsequent impact of a coral disease. Proceedings of the Fourth International Coral Reef Symposium 2:221-228. Bak R.P.M., and G. Nieuwland 1995. Long-term change in coral communities along depth gradients over leeward reefs in the Netherlands Antilles. Bulletin of Marine Science 56:609-619. Bruckner A.W., and R.J. Bruckner 1999. Rapid assessment of coral reef condition and short-term changes to corals affected by disease in Cura9ao, Netherlands Antilles. International Conference on Scientific Aspects of Coral Reef Assessment, Monitoring, and Restoration, Nova Southeastern University, Dania, FL. Abstract p. 62. Bruckner A.W., and R.J. Bruckner 2000. The presence of coral diseases on reefs surrounding Mona Island. Proceedings of the Ninth International Coral Reef Symposium Abstract, p. 281. Bythell J.C., E.H. Gladfelter, and M. Bythell 1993. Chronic and catastrophic natural mortality of three common Caribbean reef corals. Coral Reefs 12:143-152. CARICOMP 1997. Studies on Caribbean coral bleaching, 1995-1996. Proceedings of the Eighth International Coral Reef Symposium 1 :673-678. De Ruyter Van Steveninck, E.D., and R.P.M. Bak 1986. Changes in abundance of coral-bottom components related to mass mortality of the sea urchin Diadema antillarum. Marine Ecology Progress Series 34:87-94. 388 Lewis J.B. 1997. Abundance, distribution and partial mortality of the massive coral Siderastrea siderea on degrading reefs at Barbados, West Indies. Marine Pollution Bulletin 34:622-627. Meesters E.H., and R.P.M. Bak 1993. Effects of coral bleaching on tissue regeneration potential and colony survival. Marine Ecology Progress Series 96:189-198. Meesters E.H., I. Wesseling, and R.P.M. Bak 1996. Partial mortality in three species of reef-building corals and the relation with colony morphology. Bulletin of Marine Science 58:838-852. Meesters E.H., W. Pauchli, and R.P.M. Bak 1997. Predicting regeneration of physical damage on a reef-building coral by regeneration capacity and lesion shape. Marine Ecology Progress Series 146:91-99. Pors, L.P.J.J., and I. A. Nagelkerken 1998. Cura9ao, Netherlands Antilles. Pp. 127-140. In: B. Kjerfve (ed.), CARICOMP- Caribbean Coral Reef Sea Grass and Mangrove Sites Coastal Region and small island papers 3. UNESCO, Paris. Szmant, A.M. 1991. Sexual reproduction by the Caribbean reef corals Montastraea annularis and M. cavernosa. Marine Ecology Progress Series 74:13-25. Van Duyl, F.C. 1 985. Atlas of living reefs of Curagao and Bonaire. Ph.D. dissertation. Free University Amsterdam/Foundation for Scientific Research in Surinam and the Netherlands Antilles, Utrecht, 37 pp. Van Veghel, M.L.J. 1997. A field guide to the reefs of Curacao and Bonaire. Marine Ecology Progress Series 1:223-234. Van Veghel M.L.J. , and P.C. Hoetjes 1995. Effects of Tropical Storm Bret on Cura9ao reefs. Bulletin of Marine Science 56: 692-694. Weil E., and N. Knowlton 1 994. A multi-character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M. faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bulletin of Marine Science 55:151-175. Williams E.H. Jr., and L. Bunkley-Williams 1990. The world-wide coral reef bleaching cycle and related sources of coral mortality. Atoll Research Bulletin 335:1-71 Table 1. Site information for AGRRA stony coral and algal surveys off Cura9ao, Netherlands Antilles. 1998 sites are italicized. 389 390 o o3 O =3 U M— I o (U X) (U ■4—1 S i?3 O o (N c/3 13 kH O o c o cd 4^ O 'H' o o "O 'O c cd ■*-< c/3 -H C cd OJ B G O G O o TG G G 0) _N C/2 CN X G H CO) C= — •— ^ •u C3 c rf \0 ^0^00 ^ ^ CN O Tt- in fs o (N in ON (N CN ON *n fn m ON O O O Ci O O O O o c< O o o -n -H -H -H -H -+^ -H “H -H •H -H ± -H -H -H -H -H m in m C) in in in in p p in NO o O o o o o o <>S o N o 00 o CN NO r- C\ 20 cm diameter), as numbers and percentages of all colonies, by site off Cura9ao. 392 393 Plate 9A. Each stony coral, as in these colonies of Acropora cervicornis, is assessed for the presence and intensity of any bleaching that is related to mass bleaching events. (Photo Kenneth W. Marks) Plate 9B. Bleaching is characterized according to the approximate severity of tissue discoloration as pale (discoloration of coral tissues); partly bleached (patches of fully bleached or white tissue); and bleached (tissues are totally translucent, and the white skeleton is visible), as shown for this Porites. (Photo Larry Benvenuti) 394 Figure 1. AGRRA survey sites (Kalki, Jeremi, Lagun, Oostpunt) on the leeward coast of Cura9ao. CONDITION OF CORAL REEFS OFF LESS DEVELOPED COASTLINES OF CURASAO (PART 2: REEF FISHES) BY ANDREW W. BRUCKNER* and ROBIN J. BRUCKNER^ ABSTRACT Reef fish populations at 9-12 m depth in more remote eastern and western Cura9ao were characterized by low abundance, size, and diversity of large-bodied species. Families often contained only a few of the species previously reported in Cura9ao and many commercially important food species were absent. Grunts, parrotfishes, surgeonfishes and butterflyfishes were most abundant. The densities of most species averaged less than eight individuals/ 100 m and their mean sizes were 15-25 cm. Most reefs contained one or two tiger grouper (Mycteroperca tigris) and barracuda of larger sizes (-25-35 cm) and several medium- to large-sized (20-28 cm) parrotfish. No major differences were apparent between an underwater park in Oostpunt and the western reefs. The cumulative impacts of heavy fishing pressure, lack of enforcement of a spearfishing ban, along with increased urbanization and degradation of coastal nursery areas, may have contributed to the decline of Cura9ao’s reef fishes. INTRODUCTION Cura9ao, Netherlands Antilles is a small (61 km long; 443 km area) oceanic island situated 60 km north of Venezuela (Fig. 1). The island is oriented on a northwest/ southeast axis and has a narrow submarine shelf that slopes steeply into deep water. Fringing reefs surround the island but are best developed along the leeward (south) coast. The island is sparsely populated (155,000 people) with an economy based on an oil refinery and other industries, offshore banking, and tourism. There is a small commercial fishery that targets pelagic species with an estimated catch of 90-180 metric tons/year (Woodley et al., 1997). Most reef fishes for restaurant and grocery markets are imported from Venezuela although subsistence fishing is widespread along the leeward coast. 'Corresponding author, NOAA/National Marine Fisheries Service, Office of Protected Resources, 1315 East West Highway, Silver Spring, MD 20910, USA. Email: Andy.Bruckner@noaa.gov 2 NOAA/National Marine Fisheries Service, Office of Habitat Conservation, 1315 East West Highway, Silver Spring, MD 20910, USA. 396 Reef conditions and reef fish populations have degraded near population centers due to extensive coastal development, industrial activities, raw sewage discharge, sedimentation, and vessel impacts (Woodley et ah, 1997). Reefs at the eastern and western ends of the island are thought to be in better condition as they are well removed from urban and industrial development. Two small fishing villages (Lagun and Westpunt) are found at the western end of Cura9ao, but coastal development is minimal and population density is low. The Cura9ao Underwater Park extends 21 km eastward from the outskirts of Willemstad (Princess Beach Hotel) to Oostpunt and includes a 12-km stretch of coastline that is undeveloped and uninhabited. The park also contains an extensive (436 hectares) partially enclosed bay (Spaanse Water) where most of the island’s mangroves and Thalassia testudinum seagrass beds are found (Debrot et ah, 1998). Few quantitative data are available on the reef fish communities of Cura9ao (Nagelkerken, 1974; 1977; 1980). However, they are reported to be overfished as a result of heavy artisanal fishing pressure and the use of spear guns, fish traps, and gill nets (Van’t Hof et ah, 1995). The purpose of this study was to collect baseline information on the status of fish populations in more remote locations of Cura9ao at the time of initial enforcement of spearfishing prohibitions and to determine whether differences existed between the marine-park and open-access areas. The Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol was used to assess the diversity and population dynamics of reef fish with a focus on commercially and ecologically important large-bodied species. METHODS Limited estimates of the abundance and size of reef fish populations were made in four strategically chosen fringing reefs off the less populated, southeast (Oostpunt) and southwest (Jeremi, Lagun and Kalki) coasts of Cura9ao by one diver (R. Bruckner) in August 1998. Belt transects were conducted as described in the AGRRA Version 2 fish protocol (Appendix One, this volume) except that temporal constraints restricted the number of transects to three-four at each reef. All transects (30 m long, 2m X 2m window) were run parallel to depth gradients at 9-12 m depth in the same general area as the benthic surveys (Bruckner and Bruckner, this volume). Counts of serranids (groupers) were restricted to species of Epinephelus and Mycteroperca. Scarids (parrotfishes) and haemulids (grunts) less than 5 cm in total length (TL) were not tallied and only one damselfish, Microspathodon chrysurus, was recorded. Previous training dives (July 1998; n=10 dives) were conducted in Puerto Rico using styrofoam fish models (for estimating sizes) and practice transects to obtain consistency in sampling area, swimming speed, and species identification. Reef Fish Identification by Humann (1997) was used to confirm species identities. 397 RESULTS Most families of large-bodied reef fishes included in the AGRRA protocol were represented in the remote reefs of eastern and western Cura9ao although the species richness and abundance of each were relatively low at all sites (Table 1; Fig. 2). Lagun and Oostpunt had the highest numbers of species (26), while Jeremi had the highest density (57 individuals/ 100 m ). Herbivores were prominent in all reefs. Overall, parrotfishes were more common than surgeonfishes (acanthurids), but there was a notable absence of schools containing more than 6-10 adult (>10 cm TL) fish. Sparisoma viride and Scarus taeniopterus were the dominant parrotfishes; S. rubhpinne, S. aurofrenatum and S. croicensis were also common and juveniles (<10 cm TL) of these and other species were abundant in all locations. Figure 2. Mean fish abundance (no. individuals/100 ± sd) for AGRRA fishes in eastern and western Cura9ao. Other = Bodianus rufus, Microspathodon chiysurus, Sphyraena barracuda. There were no substantial differences between Oostpunt and western reefs in the overall abundance of reef fishes [pooled by family; two-factor Analysis of Variance (ANOVA); MS=0.8, F=0.95, p=0.36] although the abundance of individual families did vary (ANOVA; MS==6.6, F==8.1, p=0.004). Post-hoc analysis of major groups (pooled into herbivores, major predators, and other predators; Tukey HSD multiple comparison test) indicated that herbivores were significantly more abundant at Oostpunt with minor, non- significant differences among other groups (Fig. 2). The most abundant group of fish observed at Jeremi, Oostpunt and Lagun overall were grunts, which formed large resting 398 schools at the base of large coral heads, the dominant species being Haemulon flavolineatum and H. aurolineatum. Commercially important H. plumieri were encountered only at Oostpunt. Grunts were relatively uncommon at Kalki although three species (Anisotremus surinamensis, H. sciurus and H. flavolineatum) were observed. Oostpunt had a larger number and species of angelfishes (pomacanthids) and butterflyfishes (chaetodontids) than were recorded within transects on other reefs. No angelfish were recorded within transect areas at Jeremi whereas Chaetodon capistratus and C. sthatus were present in all four locations. Among the leatherjackets (balistids), Cantherines pullus was recorded within a belt transect only on Jeremi. Filefish (C. pullus, C. macrocerus, Aluterus schptus) were found in areas away from transects at the same depth on all reefs, and Melichthys niger commonly occurred in the water column at Oostpunt. Balistes vetula were also absent from transects in all locations but individuals were observed among gorgonians in shallow water (3-5 m depth) at Oostpunt. In addition, large schools of Caranx ruber and C. crysos were frequently sighted in the water column but not within transects. Important top predators were generally uncommon. Each reef had one-two individuals each of tiger grouper (Mycteroperca tigris) and barracuda (Sphyraena barracuda) . The most abundant snapper was Lutjanus apodus which occurred on all reefs but was most common at Lagun. Other predators recorded on most reefs at a low abundance include Epinephelus guttatus, E. cruentatus, Lutjanus apodus, L. griseus and L. mahogoni and, at Oostpunt only, Ocyurus chrysurus. The mean total length of all fishes recorded within the belt transects was generally 15-25 cm (Table 2). The largest predators were M. tigris, S. barracuda and L. apodus (25.5-35.5 cm). Large stoplight parrotfish {S. viride) were identified at Jeremi (28 cm), Oostpunt (24 cm) and Kalki (21.5 cm), but this species was smaller at Lagun (18 cm) and Size (cm) Figure 3. Size frequency distribution of herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) and carnivores (all lutjanids, select serranids) in eastern and western Curasao. 399 other species of parrotfish were mostly small (12-20.5 cm). Most grunts were small (8- 15.5 cm) with the exception of a single Haemulon macrostomum at Oostpunt (25.5 cm) and one A. surinamensis (25.5 cm) at Kalki (Table 2). The size frequency distribution of commercially important carnivores (snappers and select groupers) and key herbivores (parrotfish, surgeonfish and yellowtail damselfish) are presented in Figure 3. DISCUSSION The ecologically and commercially important species of reef fishes included in the AGRRA species list exhibited a low diversity, abundance and size in the four locations examined in Cura9ao. The dominant families seen during these surveys were Haemulidae, Scaridae and Acanthuridae, although one snapper (Lutjanus apodus) was abundant at a single location in western Cura9ao. Typically, each family was represented by one-four species with the exception of parrotfishes, for which three-six species occurred in each reef In each reef the top predators consisted of one-two tiger groupers and one-two barracudas, although these were fairly small in total length (25-35 cm). Other medium- sized fishes that were often equally or more abundant than the AGRRA-listed species included goatfish (Mullidae), squirrelfish and soldierfish (Holocentridae), and wrasses (Labridae) (R. Bruckner, unpub. obs.). From the historical data available for Cura9ao, it is apparent that many commercially important groupers, snappers, grunts, hogfish (Lachnolaimus maximus) and other species that were formerly common on these reefs (Nagelkerken, 1974, 1977, 1980, 1981) were absent or occurred at very low numbers during the present survey. While reef fish populations are known to be heavily overfished throughout Cura9ao, reefs of Oostpunt are reported to be in the best condition (Van’t Hof et al., 1995). Oostpunt’s reefs are located off an undeveloped coastline that is upstream from terrestrial sources of pollution. In addition, fish populations are offered limited protection by their inclusion in a marine park and reefs are in close proximity to important nursery areas. The mangroves, seagrass beds and shallow-reef communities in Spaanse Water were identified as the most important nursery biotope for many of the species found on fringing reefs near Oostpunt, and 32 species on the AGRRA list were identified here (Nagelkerken et al., 2001). However, there did not appear to be substantial differences between reef fish populations at Oostpunt and those at the western end of the island in terms of species composition and size. In addition, it was difficult to approach most fish in all locations with exception of small resting schools of grunts, butterflyfishes, and the small herbivores. Several factors appear to have contributed to the island-wide decline of reef fish populations. Heavy fishing pressure and the use of harmful gear (fish pots, gillnets and spearguns) are widespread and, until recently, were not managed (Van’t Hof et al., 1995). Although Oostpunt is designated as a park, the island lacks no-take marine-protected areas (MPAs) and hook-and-line fishing is permitted within park boundaries. Until 1998, spearfishing was prevalent throughout Cura9ao even though it had been prohibited more than 20 years earlier. As early as 1984, researchers reported a decline in abundance and size of groupers compared to the neighboring island of Bonaire which was thought to be 400 caused by spearfishing (Pors and Nagelkerken, 1998). The other major factor affecting fish populations may be the loss or degradation of important nursery areas. Rapid urbanization of Spaanse Water has been associated with the degradation of mangroves and seagrass beds and a precipitous decline of its coral populations (Debrot et ah, 1998). The long-term effects of these changes on reef fish populations are largely unknown. However, in a recent survey within Spaanse Water, groupers were completely absent and other commercially important species, such as white grunts (Haemulon plumieri), margates {H. album and Anisotremus surinamensis) and most snappers, occurred at very low densities (Nagelkerken et ah, 2001). This study provides only a very limited snapshot of four reefs in the more remote locations of Cura9ao and more extensive surveys are necessary to obtain a representative picture of the overall diversity, size and abundance of commercially and ecologically important reef fish. Nevertheless, these data may provide a baseline of their status at the onset of a full enforcement of the spearfishing prohibition. The elimination of spearfishing represents a critical step towards protecting and restoring Cura9ao’s reef fish populations but there is a need for other tools such as the establishment of no-take marine protected areas (Bohnsack et al., in press). It is hoped that this study will help convince resource managers and policy makers of the need for additional management measures for Cura9ao’s reefs (Van’t Hof et al., 1995). ACKNOWLEDGMENTS We gratefully acknowledge the support and facilities provided by the Caribbean Marine Biological Institute (CARMABI) and Reef Care Cura9ao. Surveys conducted at the western end of Cura9ao were supported by Hans Van den Eden of All West Divers. The opinions and views expressed in this document are those of the authors and do not necessarily reflect those of the National Marine Fisheries Service, National Oceanic and Atmospheric Administration or the U.S. Government. REFERENCES Bohnsack, J.A., B. Causey, M.P. Crosby, R.B. Griffis, M.A. Hixon, T.F. Hourigan, K.H. Koltes, J.E. Maragos, A. Simons, and J.T. Tilmant In press. A rationale for minimum 20-30% no-take protection. Ninth International Coral Reef Symposium. Bali, Indonesia. Debrot, A.O., M.M.C.E. Kuened, and K. Decker 1998. Recent declines in the coral fauna of the Spaanse Water, Cura9ao, Netherlands Antilles. Bulletin of Marine Science 63:571-580. Humann, P. 1997. Reef Fish Identification. 2"‘^ edition. New World Publications, Inc., Jacksonville, FL, 396 pp. 401 Nagelkerken, W.P. 1 974. On the occurrence of fishes in relation to corals in Cura9ao. Studies on the Fauna of Curaq^ao and other Caribbean Islands 45:1 18-141. Nagelkerken, W.P. 1977. The distribution of the graysby Petrometopon cruentatum (Lacepede) on the coral reef at the southwest coast of Cura9ao. Proceedings of the Third International Coral Reef Symposium 1:311-315. Nagelkerken, W. 1980. Coral reef fishes of Aruba, Bonaire and Cura9ao. Island Territory of Curagao. Willemstad, Cura9ao. 125 pp. Nagelkerken, W.P. 1981 . Distribution and ecology of the groupers (Serranidae) and snappers (Lutjanidae) of the Netherlands Antilles. Publication of the Foundation for Scientific Research in Surinam and the Netherlands 107:1-71. Nagelkerken, I., M. Dorenbosech, W.C.E.P. Verberk, E. Cocheret de la Moriniere, and G. van der Velde 2001. Importance of shallow- water biotopes of a Caribbean bay for juvenile coral reef fishes: patterns in biotope association, community structure and spatial distribution. Marine Ecology Progress Series 202:175-192. Pors, L.P.J. J., and LA. Nagelkerken 1998. Cura9ao, Netherlands Antilles. Pp. 127-140. In: B. Kjerfve (ed.), CARICOMP- Caribbean Coral Reef Sea Grass and Mangrove Sites Coastal Region and small island papers 3, UNESCO, Paris. Van’t Hof, T., A.O. Debrot, and I. A. Nagelkerken 1995. Cura9ao Marine Management Zone: A Plan for Sustainable Use of Cura9ao’s Reef Resources. Unpublished report, Curagao Tourism Development Bureau/CARMABI/STINAPA. 89 pp. Woodley, J.D., K. De Meyer, P. Bush, G. Ebanks-Petrie, J. Garzon-Ferreira, E. Klein, L.P.J.J. Pors, and C.M. Wilson 1997. Status of coral reefs in the south central Caribbean. Proceedings of the Eighth International Coral Reef Symposium 1:357-362. Table 1. Site information for AGRRA fish surveys in Cura9ao, Netherland Antilles. 402 E S) ' E is Q 5: S) ■ |> B .3 “ E i§ Q S: — — N Netherlands Antilles Windward Islands Saba »* St Eustatius ; Saba Bank / .J ® 15 30 miles 1 H 1 ... kilometers 0 24 48 Figure 1. Location of the windward Netherlands Antilles. A POST-HURRICANE, RAPID ASSESSMENT OF REEFS IN THE WINDWARD NETHERLANDS ANTILLES (STONY CORALS, ALGAE AND FISHES) BY KRISTI D. KLOMP* and DAVID J. KOOISTRA^ ABSTRACT Reefs of the windward Netherlands Antilles (Saba, Saba Bank, St. Eustatius, St. Maarten) were assessed at 24 sites in late 1999. The Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol was used with modifications to detect recent hurricane impacts. Live coral cover averaged 18%. The assemblage of > 10 cm stony corals was primarily composed of small-sized colonies (mean diameter » 37 cm) of which the Montastraea annularis complex was the most abundant (30% of colonies). Overall, «1% of the individually surveyed colonies had been physically damaged by Hurricane Lenny but injury levels were higher in Saba (2.6%). Bleaching was noted in >23% of colonies at the time of the assessment with the greatest percentage occurring on St. Maarten (44%) and the lowest on Saba Bank (9%). Total (recent + old) partial mortality of reef-building corals averaged less than 1 8% although levels were higher (26%) in Colpophyllia natans. Coral recruitment densities were relatively consistent (mean » 5 recruits/m ) across sites. Commercially significant fish species (i.e., serranids, lutjanids, haemulids >5 cm) were present with mean densities of 4.5 individuals/100 m . High biomass (mean « 5.8 kg/100 'y m ) of grazing, herbivorous fishes (acanthurids, scarids >5 cm, Microspathodon chrysurus) partially explains the relatively low macroalgal cover (mean » 7%) throughout this area. Saba’s fish community had a greater total biomass than those in the other three geographic areas (mean « 1 1 kg/100 m versus 7 kg/100 m ). While the coral reefs of St. Maarten show signs of disturbance (i.e., increased bleaching and sedimentation), those of Saba, Saba Bank, and southern St. Eustatius have been relatively little disturbed by coastal development and remain potential sources of marine life. Nevertheless, reef development in the windward Netherlands Antilles is limited by frequent hurricanes. INTRODUCTION The windward Netherlands Antilles (N. A.) are located at the northern arc of the leeward Lesser Antilles in the Caribbean Sea (Fig. 1). Included in the windward N. A., * Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI 48824, USA Email: klompkri@msu.edu ■y Saba National Marine Park, Fort Bay, Saba, Netherlands Antilles Email: smp@unspoiledqueen.com 406 and separated from each other by distances of 20-50 kilometers, are Saba (Fig. 2A), Saba Bank (Fig. 2B), which is totally submerged, St. Eustatius (Fig. 2C), and the southern portion of St. Martin/St. Maarten (Fig. 2D). As each represents a distinctly different area in terms of geophysical aspects and degree of human impact, they are described separately below. Saba (17°36’N, 63°15’W) is an isolated volcanic island of late Pleistocene to mid- Holocene origin (Westermann and Kiel, 1961). Saba does not rest on a carbonate shelf but a narrow submarine platform fringes the island before sloping steeply to the deep sea floor. Devoid of sandy beaches, Saba’s coastline is formed of steep, rocky cliffs. Rapid erosion of this coastline and Saba’s constricted shelf do not allow extensive reef development. Coral communities circumfuse the island occurring on granite boulders, pinnacles, and lava formations that extend seaward from the island. Close to shore, scattered coral colonies encrust large boulders that have eroded from the steep, coastal embankments. Extending seaward along Saba’s gently sloping (<10°) insular shelf are lava-flow formations aligned perpendicular to the coastline. These volcanic features, often separated from each other by sand channels, provide a basal structure for the development of coral communities. In some areas where corals have, almost contiguously encrusted lava flows, the topography superficially resembles spur-and-groove formations. Saba’s villages are located above 240 m altitude (Buchan, 1998) and coastal development is limited to a small harbor that accommodates the marine-park office, the island’s dive operators, and a recently closed rock-crushing plant. While Saba has become a popular dive destination in recent years, there appears to be minimal human impact on its reefs. (Although the authors acknowledge that Saba’s coral communities do not fit the narrow definition of coral reefs in the technical sense, we have elected to use the term throughout this paper for the sake of fluidity). However, Saba’s corals are physically damaged by hurricanes (most recently Hurricanes Luis in 1995, Georges in 1998, Lenny in 1999), especially the Acropora palmata in some shallow, high-surge zones. An area surrounding the island from the high-water mark down to a depth of 60 meters is actively regulated by the Saba National Marine Park which was established in 1987. Saba Bank (17°25’N, 63°30’W) is a submarine plateau that has been described as a sunken atoll (Vaughan, 1919; Van der Land, 1977). Its eastern edge lies about 3-5 km S W of Saba. Saba Bank is 60 to 65 km long and 30 to 40 km wide and, with a total surface area of about 2200 km, ranks among the largest atolls in the world. Rising about 1 000 meters above the surrounding sea floor, it reaches a plateau at a depth of about 1 5 m. Here aggregations of gorgonians and sparsely spaced, small (< 25 cm diameter) scleractinians are arranged linearly in tracts that are 20-40 m wide and separated by 10-20 m-wide sand channels. Depths vary between 7 and 20 m near its eastern and southeastern rim. The western ridge reportedly is deeper (50 m) and dominated by sand. From the bank’s eastern rim at «20 m depth, the bottom terrain slopes seaward to meet a terrace before it drops precipitously beyond 35 m depth. Corals are more densely spaced along 407 Figure 2. AGRRA survey sites for (A) Saba, (B) Saba Bank, (C) St. Eustatius, (D) St. Maarten, in the windward Netherlands Antilles. See Table 1 for site codes. 408 this slope (i.e., between 20-35 m depth) than at our sampling depths of 15-20 m. As depth increases, colonies become progressively larger and, for many of the Montastraea and Agaricia, increasingly flattened in morphology. Van der Land (1977), who referred to this coral-covered slope as the “front reef,” noted that Saba Bank was an island during the last glacial period and until at least about 5,000 years ago. He hypothesized that this windward “front reef’ originated, when sea levels were 20-30 m below present, as a fringing reef of Saba Bank's island. Because a considerable distance separates Saba Bank from land, human impact on its corals is restricted to that imposed by fishing activity, boats and ships. Part of the Bank is within the geographic scope of Saba island’s jurisdiction as territorial sea; the larger part falls within the limits of a possible Exclusive Economic Zone (EEZ) within the competency of the Netherlands Antilles. At present, only a limited EEZ has been claimed by the Netherlands Antilles in the form of an Economic Fisheries Zone. Van der Land (1977), Meesters et al. (1996) and Dilrosun (1999) have ascertained that, in isolated areas, Saba Bank’s reefs are rich in terms of cover and diversity of reef-building corals but that its fish stocks appeared to be declining. However, its reefs have remained largely unexplored. Hence, Saba Bank provides us with a rare opportunity to study the regional potential for reef development in the absence of coastal processes. Sint Eustatius (17°29’N, 62°57’W) is situated on a submerged platform shared with the islands of St. Kitts and Nevis. The island consists of two extinct volcanoes, an older (late Pliocene) eroded volcano in the northwest and a younger volcano with evidence of activity as recent as the early Pleistocene epoch (Adey and Burke, 1976) in the south. A narrow platform (<2 km wide) surrounds most of the island. Fringing reefs occur mid-shelf off its southern and northwestern coasts. The remainder of the shelf is a flat, sandy plateau with limited potential for coral development. Lava flow formations from the southern volcano extend seaward and, along with large volcanic boulders, provide a high-relief (»3-4 m) structural base for reefs on the island’s southern side. These reef structures, each about 30-50 m wide, are separated by sand channels of !«50 m width and frequently occur in a series. St. Eustatius' most robust reefs in terms of coral cover occur mid-shelf near the island's southernmost tip. Here the underlying substrata on which the corals occur is carbonate in nature but it is not apparent from casual observation whether these are entirely of carbonate composition or whether they embody inorganic, pyroclastic fragments. While typically spanning an area similar in size (30-100 m in any direction), they are not as eonsolidated as lava flows nor do they exhibit the uniform orientation ("seaward") of lava flow. However, these formations are equal in vertical relief (rising 3-4 m from the sand-covered bottom) and, where disjunct, are often cemented together with mature (>100 cm diameter) colonies of Montastraea. Corals also fringe the rocky northwestern coast near shore where broken skeletons of A. palmata and live stands of Millepora occur on boulders in the high-surge zones. Further from shore, scattered coral heads occur along a reef flat, though most are dead and covered with sediment. The coastline of St. Eustatius is relatively undeveloped with the exception of its capital city, Oranjestad, and a trans-shipment oil terminal, both located on the western coast. The St. Eustatius Marine Park was established in January 1998 and is operated by the St. Eustatius National Park Foundation (STENAPA). Two areas, the Northern Marine Park and the Southern Marine Park, are actively protected by St. Eustatius Marine Park 409 officials and fishing activity is restricted in these zones. St. Eustatius’ reefs have been impacted by hurricanes in recent years and those in the Northern Park have been severely impacted by anthropogenically enhanced sedimentation. Saint Martin/Saint Maarten (18°N, 63°W) is situated on the Anguilla Bank, a microplate shared by the islands of Anguilla and St. Barthelemy. The Anguilla Bank is comprised of an old (early to mid-Tertiary period) basal rock of fine mudstones and the islands represent an uplifted veneer of young (Pleistocene to Recent) reef limestones and sandstones (Oxenford and Hunte, 1990). A fringing reef flat occurs mid-shelf («2-3 km) off the southern coast of the island and fringing reefs are found adjacent to small islets located at similar distances from St. Maarten’s southeastern coast. The reefs off the southern coast emerge from the surrounding hardbottom at depths of about 8-15 m. Most of this system is characterized by low-relief, scattered corals but a spur-and-groove system of intermediate relief (<2 m) occupies a small area. These reefs are routinely exposed to strong offshore currents. Remnants of Acropora palmata are apparent within the spur-and-groove zone but most colonies are either “standing dead” (i.e., entirely dead but still attached to the substratum) or, if still living, have spread horizontally without having attained much height above the substratum. The reefs which fringe the small islets on St. Maarten's southeastern side grow compactly along a slope (<30°) with no lagoonal area separating them from the islets. Tourism is a primary contributor to St. Maarten’s economy. St. Maarten is a major port of call for Caribbean cruise ships and, at the time of this survey, operations were under way to expand harbor capacity to accommodate additional vessels. As the island’s infrastructure struggles to meet the demands of development, St. Maarten’s reefs are increasingly threatened by devegetation, siltation, sewage input, recreational boating (Nijkamp et al., 1995), and anchor damage (Smith et al., 1997). Additionally, Smith et al. (1997) reported that heavy seas during Hurricane Luis (1995) resulted in resuspended sand, smothering reefs and damaging A. palmata in shallow water. The Marine Park of St. Maarten, established in 1997 in conjunction with the Nature Foundation St. Maarten, encompasses a portion of its reefs. At the time of this assessment, the marine park was in its initial stages of devising a management plan. One of our survey sites (SXMOl), which lies within the marine park boundaries (Fig. 2D), has been the site of two cruise ship groundings in recent years. The windward N. A. are frequently impacted by hurricanes. During the previous decade, a succession of severe storms (Hugo in 1989, Luis and Marilyn in 1995, Bertha in 1996, Georges in 1998, Jose and Lenny in 1999) had came near enough to impose harm to the reefs. Prior to Hurricane Luis in 1995, Acropora palmata was the predominant scleractinian on many of Saba’s shallow (<10 m) reefs (i.e., at sites 1, 2, 8, 9; Fig. 2A). Although some recovery of A. palmata was evident following Hurricane Luis, they sustained further damage during Hurricanes Georges (1998) and Lenny (1999). Hence, at many Saban sites, A. palmata has been replaced by gorgonians and Millepora spp. (Kooista, personal observations). Our surveys were conducted less than two weeks after the passage of Hurricane Lenny, a category 4 (Safir/Simpson scale) hurricane, which coursed between St. Maarten and Saba on November 18 and 19, 1999. Coupled with the impact of Hurricane Jose (category 2), which had earlier passed through the windward N. A. islands in October 410 1999, we were provided with an opportunity to assess damages associated with recent hurricanes. METHODS The surveys were conducted in the windward N. A. (Fig. 1) during November and December 1999. Data were collected from nine sites along the coast of Saba (Fig. 2A), three sites along the eastern rim of Saba Bank (Fig. 2B), 10 sites along St. Eustatius’ southwestern shore (Fig. 2C), and two sites along St. Maarten’s southern and southeastern shore (Fig. 2D). These sites were among those chosen during a reconnaissance evaluation in October 1999 to be representative of reefs within the area and to strategically accommodate marine park management objectives. Due to reduced visibility associated with storm-induced rainfall and runoff, we omitted sampling where conditions were prohibitive including the reefs in St. Eustatius’ Northern Marine Park. Sites were surveyed using the AGRRA protocols version 2.1 (see Appendix One, (this volume). The benthic surveys were made by four divers. In quantifying individual stony corals, we modified the method to include all colonies >10 cm in maximum diameter to ensure adequate sample sizes. Coral sizes were measured to the nearest cm. Additionally, we recorded incidence of recent hurricane damage along each benthic transect as indicated by overturned, broken, or abraded coral colonies. Stony corals were searched for damselfish algal gardens and the numbers of benthic damselfish along the transect line were also counted. The Montastraea annularis complex was treated as a single species with three morphologically distinct formae. Sediment in the algal quadrats was wafted away by hand before scoring the relative abundance of crustose coralline algae. Macroalgal heights were measured to the nearest cm. Two divers conducted all the fish surveys between 9:00 a.m. and 4:00 p.m. Counts of serranids (groupers) were restricted to species of Epinephelus and Mycteroperca; scarids (parrotfishes) and haemulids (grunts) less than 5 cm in length were not tallied. Roving Diver observations averaged 30-45 minutes each. Fish biomass was estimated using length-weight relationships given in Appendix Two (this volume). Species recorded during the roving dives were assigned to trophic guilds developed by Schmitt et al. (1998) based on the major source of food in their diet determined from prior studies (Randall, 1967). Field guides included Humann (1992, 1993, 1994), Goodson (1976), Littler et al. (1989), and Smith (1948). Parameters were statistically analyzed to detect differences between the four geographical areas using Statistica software. Version 5.1 (StatSoft, Inc., 1998). To reduce heteroscedasticity of variance, a single-factor, Kruskal-Wallis nonparametric Analysis of Variance (ANOVA) by rank was used. Kruskal-Wallis nonparametric ANOVA by rank was also used to detect differences among sites within each of the four geographical areas. The Tukey HSD for unequal N was used, when relevant, as an ad hoc multiple comparison test. Significance was tested at alpha = 0.05. Coral species composition and coral size distribution were analyzed using a chi-square analysis to detect proportional differences among the four geographical areas. Where no significant differences were found, the data were combined as a single value for the windward N, A. Fish densities and biomass were calculated for sites within the four areas and collectively for the 411 windward N. A. A linear regression was performed to examine the relationship between the biomass of herbivorous fishes (i.e., acanthurids, scarids >5 cm, Microspathodon chrysurus) and macroalgal abundance. RESULTS Stony Corals at the Reef Sites The 284 benthic transects were made in depths of 6-20 m at the 24 survey sites (Table 1). Mean live coral cover for the windward N. A. (Fig. 3) was 18% (se = 5.7). Though not significantly different from the other three geographical areas (ANOVA,/? = 0.33), the lowest live coral cover (1 1% ± 2.4 se) was found in St. Maarten. Significant, between-site differences in live coral cover (Table 1) were evident in the three remaining areas (for each ANOVA,/?<0.001, Tukey p <0.05). Sites SAB03 and SAB09 in Saba (mean = 8% ± 1 .1 se) were significantly lower than the other Saban sites (mean = 20% ± 3.9 se), but comparable in live coral coverage to the St. Maarten sites (ANOVA,/? = 0.12). All three sites on Saba Bank differed significantly from each other. Sites EUXOl through EUX08 in St. Eustatius were statistically similar (mean = 16% ± 1.6 se) and substantially lower than EUX09 and EUXIO (mean = 46% ± 1.5 se). 45 n 40 - 35 J 30 J 25 ] 20 -i f ! 15 j 10 -L ^ i 5 H 0 ^ — L_ . — — , — — , SABA SABA ST. ST. BANK EUSTATIUS MAARTEN n=9 n=3 n=10 Figure 3. Mean live stony coral cover (percent ± se) for each geographical area in the windward Netherlands Antilles. Dotted line represents overall mean. n=number of sites. 0) > o O ‘to o O 0) > The 2,930 colonies of stony corals, each at least 10 cm in diameter, that were encountered along the transect lines represented 24 species of stony corals (23 scleractinians and the hydrozoan Millepora complanata). Montastraea was the most abundant stony coral genus throughout the windward N. A. (40% of all individually surveyed colonies). Montastraea annularis faveolata was the predominant taxon (19% 412 A Windward Netherlands Antilles Total number of colonies surveyed = 2930 AA OTHER CN B Saba Colonies surveyed =1155 Taxon Code AA = Agaricia agaricites (n=245) CN = Colpophyllia natans (n=l 1 1) DL = Diploria labyrinthiformis (n=155) DS = Diploria strigosa (n=399) MAP = Montastraea annularis faveolata (n=525) MFR = Montastraea annularis franksi (n=3 19) MC = Montastraea cavernosa (n=292) MIL = Millepora complanata (n=55) PA = Porites astreoides (n=348) SS = Siderastrea siderea (n=161) C St. Eustatius Colonies surveyed = 1230 MC MFR 10% 10% D Saba Bank Colonies surveyed = 323 23% OTHER aA 12% 170/^ E St. Maarten Colonies surveyed = 222 Figure 4. Species composition and mean relative abundance of all stony corals (>10 cm diameter) in (A) windward Netherlands Antilles, (B) Saba, (C) St. Eustatius, (D) Saba Bank, (E) St. Maarten. 413 of all colonies), and nine species accounted for nearly 90% of the “major reef-building corals” (Fig. 4). Acropora palmata, most of which were located near Green Island at SAB09, comprised <1% of these corals. Only two colonies of A. cervicornis were encountered during our entire survey. Results of a chi-square analysis for proportional similarity suggest that the individually surveyed corals in Saba and St. Eustatius are statistically similar in species composition (chi-square, p = 0.35). Likewise, those of Saba Bank and St. Eustatius show similar proportions (chi-square, /? = 0.31). Marginally significant similarities in species composition were detected between Saba and St. Maarten (chi-square, = 0.1 1). Significant similarities were lacking for the three remaining geographic comparisons (each chi-square, p <0.015). Coral recruits were observed at all sites with an overall mean density equivalent to 4.7 recruits/m (Table 3). Recruitment levels were not significantly different between the four areas (ANOVA,/? = 0.10) and, although variable between transects, were not significantly different among sites at each of the four geographical areas (each ANOVA, p > 0.12). Recruits were composed primarily of Siderastrea spp., P. astreoides, and Agaricia agaricites (Fig. 5). Condition of Individually Surveyed Corals (at least 10 cm in Diameter) On average, the maximum diameter of the > 10 cm corals in the windward N. A. was 37 cm and their height was 19 cm. In Saba, however, many colonies of Diploria spp. and P. astreoides form thin crusts (<10 cm high) as they spread over boulders that have eroded from the island. Mean colony size was notably greater (Table 2) at three sites (BNKOl, EUX09, EUXIO) where the species composition was more heavily dominated (44%, 51%, and 43 %, respectively) by Montastraea annularis faveolata. In addition, mean diameters of Montastraea annularis faveolata were greater at these three sites ( 1 05 cm, 122 cm, 122 cm) than was documented at other sites (M annularis faveolata, mean diameter; BNK = 95 cm; EUX = 83 cm). Size distributions for the five most common coral taxa (Fig. 6) reveal that M. annularis faveolata had the largest colonies (maximum diameter = 610 cm). The remaining corals were considerably smaller, especially the two smallest {P. astreoides, A. agaricites). Sample sizes were adequate to perform a chi-square analysis of size distribution among three geographical areas for four of the common coral taxa. Significant differences in size distribution were only noted in M. annularis franksi (Fig. 7), for which a higher proportion of smaller colonies (< 30 cm diameter) was found at St. Eustatius than in Saba and the Saba Bank {p < 0.001, both locations). The impact of Hurricane Lenny was evident wherever coral colonies or fragments had been freshly dislodged. Recent storm damage was greatest at Saba, particularly at sites 4, 8 and 9 (where 4.5%, 6.5%, and 4.5%, respectively, of the individually surveyed stony corals were affected), but averaged about 1% for the windward N. A. as a whole (Table 2). It is noteworthy to mention that Millepora complanata, although it only represented «2% of the surveyed corals, constituted about 40% of the colonies that were damaged in the storm. While physical damage to corals was less severe in St. Maarten than Saba (Table 2), it was apparent that large quantities of sediment had been transported by currents here, burying some structures and exposing others. 414 A B D Windward Netherlands Antilles Other AA 1243 quadrats Taxon Code AA=Agaricia agaricites ?A=Porites astreoides S,K=Siderastrea radians SS=Siderastrea siderea Other Acropora cervicornis Dichocoenia stokesii Diploria strigosa Montastraea cavernosa Madracis dec act is Montastraea annularis faveolata Millepora Mussa angulosa Porites porites Stephanocoenia intersepta Unknown Saba Other 14% AA 9% n=139 484 quadrats 44% C St. Eustatius Other aa 494 quadrats 38% 5% PA 23% Saba Bank E St. Maarten AA 7% Figure 5. Species composition and mean relative abundance of all stony coral recruits (<2 cm diameter) in (A) windward Netherlands Antilles, (B) Saba, (C) Saba Bank, (D) St. Eustatius, (E) St. Maarten. 415 Montastraea annularis faveolata (18% of colonies) Montastraea annularis franks! (11% of colonies) 30 H n=319 Diploria strigosa (14% of colonies) Porites astreoides (12% of colonies) Montastraea cavernosa (10% of colonies) 70 60 ‘ n=292 A CO cn cn V M O o o o cp O 1 ro 1 CO 1 1 cn cn o CD CO CO CD CO Agaricia agaricites (8% of colonies) A OJ cn cn V N) O o o o o -vj O 1 N) 1 Oc> 1 1 cn 1 cn o CO CD CD CO CD Colony Diameter (cm) Figure 6. Size-frequency distribution of all colonies (>10 cm diameter) of Montastraea annularis faveolata, M. annularis franksi, Diploria strigosa, Porites astreoides, Montastraea cavernosa, Agaricia agaricites in the windward Netherlands Antilles. 416 CO c o ‘-4-^ (0 0) W XI o >» o c 0) :3 cr 0) A Saba A N) Ca3 4:^ cn (J) 00 CD V O O O o

^>1 00 CD V n: cp o O o o O O O o o CO cn dn do CD o

10 cm diameter) for (A) Saba, (B) Saba Bank, (C) St. Eustatius. St. Maarten was not included in the analysis due to insufficient sample size. 417 Over 23% of the > 10 cm stony corals exhibited some degree of bleaching (Table 2) but only 1% of these colonies were totally bleached. Of the bleached colonies throughout the N. A., more colonies were partly bleached (12%) than uniformly pale (9%). Bleaching was most prevalent in St. Maarten (44% of all colonies) where the percentage of pale colonies exceeded that found in the other three geographical areas (SXM = 25%; SAB == 10%; BNK = 2%; EUX = 9%). The Saba Bank sites were the least bleached (9% of all colonies). Intermediate levels of coral bleaching were found at Saba and St. Eustatius (22% and 24%, respectively). Diseases were noted in about 0.7% of the individually surveyed corals and showed relatively little geographic variation in abundance (Table 2). Yellow-blotch (yellow-band) disease, primarily in the Montastraea annularis complex, was the most prevalent disease affecting about 0.4% of the colonies overall. Black-band disease, occurring in 0.1% of the stony corals, affected Diploria strigosa, Agaricia agaricites, M. annularis faveolata, and M. cavernosa. Darkspots disease occurred in 0.1% of the colonies and was only seen in Siderastrea siderea. White-band disease was seen in Acropora palmata and affected 0.1% of the stony corals. White-plague disease was rare and seen in < 0.1% of the colonies of Colpophyllia natans. Figure 8. Frequency distribution of total (recent and old) partial colony mortality of all stony corals (>10 cm diameter) in the windward Netherlands Antilles. Partial-colony mortality of the individually surveyed corals averaged 1.0% for recent mortality and 16% for old mortality (Table 2). Forty-seven percent of these colonies showed no evidence of any kind of mortality and most of the remainder had less than 2% recent and 30% total (recent + old) partial mortality (Fig. 8). Although there were no significant differences among the four areas in recent, old, or total partial mortality (K-W ANOVA,/? > 0.35), partial mortality varied by taxon (Fig. 9). Porites 418 astreoides exhibited the lowest levels of recent and old partial mortality (0.5% and 5.0%, respectively). Recent mortality was highest in C. natans (3.5%) and Diploria strigosa (3.0%). Old (and total) mortality was also substantially higher in C. natans (26%). Overall, 0.7% of all surveyed corals were standing dead (Table 2). However, 45% of the A. palmata colonies we surveyed were determined to be standing dead and their skeletons represented 56% of the total standing dead in the windward N. A. survey. ■e o Taxon Code CN= Colpophyllia natans SS=Siderastrea siderea DS=Diplona strigosa MC=Montastraea cavernosa MFR=Montastraea annularis franksi DL=Diploria labyrinthiformis y[.AF=Montastraea annularis faveolata MlL=Millepora complanata AA=Agaricia agaricites PA= Porites astreoides CN SS DS MC MFR DL MAF MIL AA PA Coral Species Figure 9. Mean percent total (recent and old) partial colony mortality of common stony corals (>10 cm diameter) in the windward Netherlands Antilles. Algal Groups, Diadema, and Damselfishes Relative algal coverage was assessed for a total of 1 ,242 quadrats in the windward N. A. Turf algae were the predominant functional group at 13 sites, and were codominants with crustose coralline algae at a further seven sites. Turf relative abundance overall averaged 53%, while the crustose corallines and macroalgae averaged 42% and 7%, respectively. Where macroalgae were present, they displayed a low vertical profile (<2 cm at 21/24 sites) so that macroalgal indices (% relative abundance of macroalgae x macroalgal height-a proxy for macroalgal biomass) everywhere were <40 (Table 3). Most of the macroalgae we encountered were fleshy species, mainly Dictyota, but some calcareous Halimeda were also present. No Diadema antillarum were found in the benthic belt transects and Diadema were only sighted twice during any of our dives. 419 Benthic damselfish were detected in quantities of «1.5 individuals per 10 m transect (Table 2) and were represented primarily by the yellowtail damselfish, Microspathodon chrysurus. The threespot damselfish, Stegastes partitus, was rare as were its algal gardens in the surveyed corals. Fish Diversity A total of 142 species of fish were recorded during 25.5 hours of roving diver surveys conducted at the 24 sites of the windward N. A. The fish assemblages were represented by six major trophic guilds with herbivores being dominant (28% of species sighted) and species from planktivorous and piscivorous guilds each representing 24% of the sightings (Table 4; Fig. 10). Sightings of serranids included Epinephelus fulvus (coney, 100% of dives), E. cruentatus (graysby, 86% of dives), E. guttatus (red hind, 52% of dives), Mycteroperca venenosa (yellowfm grouper, 24% of dives), M. tigris (tiger grouper, 21% of dives), E. striatus (Nassau grouper, 7% of dives), M interstitialis (yellowmouth grouper, 3% of dives), and E. morio (red grouper, 3% of dives). Although no fish kills were noted, many of the fish, especially in St. Maarten and Saba, bore bruises and lacerations presumably resulting from Hurricane Lenny. These were particularly apparent on the pale, smooth skin of acanthurids but also noted on many scarids. Additionally, it appeared that a greater number of fish than normal were spending time at cleaning stations, seemingly having their injuries tended. Figure 10. Trophic guild composition of the 25 most commonly sighted fish species during roving diver surveys in the windward Netherlands Antilles. 420 Figure 11. Mean fish density (no. individuals/1 OOm^ ± se) for AGRRA fishes in the windward Netherlands Antilles. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus and Sphyraena barracuda. Ecologically and Commercially Significant Fishes A total of 6,1 16 individual fish were counted in 240 belt transects. Acanthurids (surgeonfish) and scarids dominated the selected species surveyed in the belt transects (Fig. 1 1). The most common species were Acanthurus bahianus, A. coeruleus, Sparisoma aurofrenatum, S. viride, Scarus croicensis and S. taeniopterus. Although variable among sites and areas, acanthurids had the greatest densities (mean = 19 individuals/ 100 m^) and highest biomass (mean = 3,150 grams/100 m^) overall (Table 5). Total lengths for four common species of parrotfish (counts restricted to >5 cm) averaged 19 cm for Scarus croicensis, 23 cm for S. taeniopterus, 20 cm for Sparisoma aurofrenatum and 28 cm for S. viride. No significant differences were detected in mean lengths among the four geographical areas for S. croicensis (ANOVA,/? = 0.20) and for S. viride (ANOVA,p = 0.51). However, S. aurofrenatum were significantly larger in Saba (ANOVA,p = 0.01) than on the Saba Bank (Tukey,/? = 0.05) and in St. Eustatius (Tukey,p = 0.01), but overlapped with those in St. Maarten where mean lengths were highly variable (Fig. 12). Statistical comparisons for S. taeniopterus were not computed since this species was abundant only in St. Eustatius. The density of herbivorous fishes (i.e., acanthurids, scarids >5 cm, M. chrysurus) 'y averaged 30.8 individuals/ 100 m . Acanthurids were predominant with an overall mean density of 19.3 individuals/100 m^ followed by scarids and M. chrysurus with respective means of 9.3 and 2.2 individuals/100 m^. This pattern of dominance was consistent throughout the four geographical areas surveyed with the exception of Saba Bank where densities of acanthurids and scarids were nearly equal. However, mean density of the SABA SABA ST. ST. BANK EUSTATIUS MAARTEN n=72 n=54 n=79 n=2 SABA n=14 SABA ST. ST. BANK EUSTATIUS MAARTEN n=82 n=27 n=56 SABA SABA ST. ST. BANK EUSTATIUS MAARTEN n=61 n=33 n=56 n=32 Figure 12. Mean total length (± se) for three common scarids (all >5 cm) for each geographical area the windward Netherlands Antilles. 422 80 - 60 - 40 J 20 - 0 - Scaridae n=1345 0-5 6-10 11-20 21-30 31-40 >40 0-5 6-10 11-20 21-30 31-40 >40 Size Category (cm) Figure 13. Size frequency distribution of dominant herbivores in the windward Netherlands Antilles. Acanthuridae was notably greater in Saba than in the other three areas (Table 5). Most of the herbivorous fish we recorded were between 1 1 and 30 cm in length (Fig. 13). The mean density of fish species of high commercial value (i.e., large serranids, lutjanids, haemulids >5 cm) was 4.5 individuals/ 100 m . The Serranidae, of which the numerical dominant was Epinephelus fulvus (with 85% of individuals), was the predominant family with an overall mean of 2.4 individuals/ 100 m^ (Table 5). The most common size class for individual serranids and lutjanids in the windward N. A. was 21- 30 cm whereas haemulids were evenly distributed between 1 1-20 cm and 21-30 cm (Fig. 14). Total community biomass for the eight fish families included in the AGRRA belt transects averaged 16 kg/100 m^ (± 5.1 se). Saba (Fig. 15) had a significantly greater biomass (mean = 1 1 kg/ 100 m^ ± 3.5 se; /?=0.0003) than the other geographical areas (combined mean = 7 kg/100 m^ ± 2.0 se). 423 (/) c g OJ 0) (/) x> O >« o c (U 3 cr 0) 60 n 40 ^ 20 - 0 - 0-5 6-10 11-20 21-30 31-40 >40 60 n 40 - 20 - 0 - Haemulidae 0-5 6-10 11-20 21-30 31-40 >40 Size Category (cm) Figure 14. Size frequency distribution of commercially valuable carnivores in the windward Netherlands Antilles. SABA SABA BANK ST. ST. EUSTATIUS MAARTEN Figure 15. Total mean biomass (kg/100 m^± se) for the AGRRA fishes in eight families in the windward Netherlands Antilles. 424 A linear regression revealed a significant inverse relationship {p = 0.03; r^ =0.19) between herbivorous fish biomass and macroalgal index at the windward N. A. survey sites (Fig. 16). Figure 16. Regression plot between mean herbivore biomass (kg/lOOm^) and mean macroalgal index, by site in the windward Netherlands Antilles. DISCUSSION The condition of reefs in the windward N. A. is largely influenced by their position within the hurricane belt of the Atlantic Ocean. The vulnerability of this region to the impacts of storm forces is evidenced by the disappearance of mature stands of Acropora palmata in recent years (Acroporids are particularly susceptible to damaging storm forces due to their fragile, branching nature.). Before Hurricane Luis (1995) A. palmata dominated sites SABI and SAB2 on Saba’s windswept east coast, along with sites SABS and SAB9 on the north coast which also frequently experience high waves and strong surges (ENCAMP, 1980a). Although some recovery had been observed since Luis (Kooistra, personal observations), Hurricanes Georges (1998) and Lenny (1999) caused still more damage. We were unable to detect most of the colonies of A. palmata which died during earlier storms as the remaining skeletons have been eroded beyond taxon recognition. The overall small size of the individually surveyed corals at most sites in the windward N. A. (Table 2; Fig. 6) reflects the high abundance of small colonies of Diploria spp., Porites astreoides and Agaricia agaricites. The flattened growth patterns of Diploria spp. and P. astreoides on many Saban reefs may reflect an adaptation to high surge and frequent hurricanes. 425 Hurricanes normally develop over the Atlantic and approach the windward N. A. islands from the east. Hurricane Lenny was unusual in that it developed in the southern Caribbean and advanced from west to east. While situated between St. Maarten and Saba, the hurricane stalled in position for nearly 22 hours which may explain why the greatest physical damage to stony corals was experienced in the Saban sites with western exposure (SAB4, SABS, SAB9). During our dives, visibility was somewhat restricted (ranging from 4-23 meters) due to increased turbidity from suspended sediments following Hurricane Lenny’s passage. The high levels of partial bleaching that we recorded (nearly 25% of the individually surveyed colonies) was likely a response to conditions of reduced light due to the increased turbidity. Bleaehing conditions may have been compounded by circumstances (i.e., suspended sediment and turbidity) introduced by Hurricane Jose in October 1999, although we have no evidence to support this claim. Although some settling of sediment had occurred at the time of our assessment, the full extent of the hurricane’s impact on the reefs of the windward N. A. might not be reeognizable since delayed mortality due to sedimentation will be difficult to attribute to this storm with any degree of certainty. Additional damage was noted in large (up to 1 m high) barrel sponges. Some of the Xestospongia muta, which were previously abundant on the reefs of Saba and St. Eustatius, disappeared during Hurricane Lenny while others were severely damaged. Regrowth rates of approximately 2.5 cm per month were initially recorded in the surviving sponges (Kooistra, personal observation) although recovery rates slowed after several months. It remains unclear why C. natans displayed a higher incidence of partial mortality (particularly old mortality) than these other species. Incidence of disease in C. natans and in Diploria strigosa (for which recent partial mortality was also relatively elevated) were no higher than in other taxa. Other potentially aggravating agents such as Corallophilia abbreviata, Palythoa spp., and burrowing serpulids were noted but C. natans and D. strigosa were not disproportionately plagued by these factors. Partial mortality in the > 10 cm corals was low throughout the windward N. A. in late 1999 (Fig. 8). It should be recognized, however, that all of our sites were situated in deeper water (by AGRRA standards) where mortality is normally less prominent. In all likelihood, mortality estimates would have been greater had we been able to document the demise of the Acropora that once existed in these waters (see above). The Montastraea annularis complex consists of slow-growing massive corals and, like the acroporids, has been an important primary reef-builder in the Caribbean. If future climate conditions resemble contemporary patterns, it is unlikely that the reefs of the windward N. A. will sustain substantial populations oi Acropora. Hence, the importance of the massive Montastraea annularis complex as a reef-building component is reinforced as it is more resilient to destructive storm forces. Montastraea has been a persistent genus on reefs over geologic time (Birkeland, 1997). Westerman and Kiel (1961) reported “plate-like” eolonies of fossilized M. annularis embedded in St. Eustatius’ coastal cliffs (Sugar Loaf) near our sites EUX09 and EUXIO. Its presence in this region 20,000 to 70,000 years ago, and as mature living colonies today, suggests that local environmental conditions have been suitable and recruitment successful for this species throughout glacial sea-level oscillations and disturbances during the Recent. The 426 potential exists for reef development throughout this area if environmental conditions remain favorable and if M annularis continues to proliferate and withstand disturbances. Notwithstanding its dominance throughout the windward N. A. (the Montastraea annularis complex collectively constituted 30% of the individually surveyed corals), none of the various morphs had been a major contributor (< 1%) to coral recruitment (Fig. 5) in late 1999. Our findings are consistent, however, with previous studies of coral recruitment within the Caribbean. Rylaarsdam (1983) reported that Montastraea annularis composed < 1% of juveniles recorded at Discovery Bay, Jamaica where its colonies made up 7-32% of the coral cover. Hughes (1985) noted complete failure to detect M annularis recruitment on reefs at Rio Bueno, Jamaica where larger colonies were locally abundant. Similarly, a paucity of juveniles of reef-building taxa from Curacao and Bonaire led Bak and Engel (1979) to conclude that the composition of the adult coral community is not a direct function of juvenile abundance. The low rate of recruitment of M. annularis on Caribbean reefs has been explained as a characteristic of its life history strategy. This is in contrast to species which commonly dominate the juvenile community and that display relatively high rates of recruitment followed by high mortality (eg., Agaricia agaricites and Porites astreoides). M. annularis has a relatively high survival rate aided by its moderate ability to clear sediment from its surface, its aggressive nature toward other species, and its ability to regenerate lesions (Bak and Engel, 1979; Wittenberg and Hunte, 1992). However, since it is unknown what threshold level of parental spawning stock is necessary to maintain coral populations at sustainable levels, a decrease in adult spawners, as evidenced by a decline in live coral cover, may indicate a reduced capacity to replenish populations. Long-term monitoring studies in the U.S. Virgin Islands where M. annularis is the dominant coral indicated that following a 35% decline in its cover, there had been no substantial recovery (Smith et al., 1997). Subsequent recruitment studies in this area revealed that broadcast spawning species (including the M annularis complex and Acropora spp.) composed less than 4% of coral recruits (Kojis and Quinn, 2001). Even if relatively low densities of juvenile M annularis are a natural phenomenon, these levels may not be adequate to offset reduced reproduction associated with high rates of decline in live coral cover. Hence, the importance of M annularis as a major component to reef structure in the N. A. may warrant monitoring to detect future trends. The rather high relative abundance of crustose coralline algae (Table 3) is suggestive of conditions that should be conducive for the settlement and recruitment of coral larvae. Indeed, while recruitment was highly variable, corals less than 2 cm in diameter were detected at all sites. A low standing stock of macroalgae and low macroalgal indices (Table 3) may indicate a seasonal response to reduced light and water temperatures (Littler et al., 1986). Alternatively, it may imply that algae are routinely nutrient-limited or that algal biomass is kept in check by grazing herbivores. There are no data to support nutrient limitation or enrichment and the grazing echinoid, D. antillarum, was not present in our transects. Herbivorous fishes, especially acanthurids, were present in high densities at most sites (Table 5), however, and macroalgal biomass was inversely related to the biomass of herbivorous fishes in late 1999 (Fig. 16). Unfortunately, we have no quantitative data by which to compare algal conditions prior to Hurricane Lenny with those at the time of our assessment. Fleshy macroalgae (e.g., Dictyotd) were abundant at two sites in St. Eustatius (EUX03, EUX05) during reconnaissance dives in 427 October 1999 but were not conspicuous elsewhere at that time. Considering that large barrel sponges were displaced by Hurricane Lenny, it is reasonable to believe that many of the macroalgae would also have been dislodged by this storm. The fish assemblage of the windward N. A. is fairly rich in species and well represented by trophic guilds (Fig. 10). A relatively high density of herbivorous acanthurids and scarids was present at the time of our assessment (Table 5; Fig. 1 1). That we found greater mean total lengths for Sparisoma aurofrenatum in Saba (and possibly St. Maarten) than at Saba Bank and St. Eustatius (Fig. 12) may be an indication of better growth conditions at the former or it could be related to variations in age distribution among the geographic areas. Since these species are not targeted by the fishing community, small deviations in size are not of great concern but may be a point of interest to the dive community, or for future growth or size-related studies. A number of studies have demonstrated that macroalgae appear to be common only where they are not heavily grazed (Lewis, 1986; Morrison, 1988; Schmitt, 1997). The inverse relationship we detected between herbivorous fish biomass and macroalgal biomass (Fig. 16) may suggest that herbivorous fish biomass increased as a result of effective grazing although this relationship only explains 19% of the variation in herbivore biomass. Even though most reef fish are generally considered “nonmigratory,” foraging groups of scarids and acanthurids are known to cover distances of hundreds of meters to several kilometers from their “home” reef (Reeson, 1983; Risk, 1998; Williams, 1991). Since fish biomass is a consequence of preceding food conditions and may have been attained at a time and locale where algal conditions varied from what we sampled in 1999, caution must be exercised in drawing conclusions based on a “snapshof ’ algal/fish biomass relationship. It is noteworthy to recognize that most of the serranids encountered in our belt transects were Epinepheus fulvus (coney), a smaller (reaching « 30 cm) less-targeted species. Since many of the larger serranids are wary of divers, it is possible that the belt- transect method did not adequately quantify their abundance. We did not detect many individuals of commercial value less than 10 cm (Fig. 14) which is not unexpected since our assessment took place in relatively deep water and it is recognized that juvenile serranids, lutjanids, and haemulids occupy shallow- water habitats (Thompson and Munro, 1983a; Thompson and Mumo, 1983b; Gaut and Munro, 1983). Although lutjanids and serranids are known to feed on a wide range of prey, they are primarily piscivorous (Parrish, 1987). Their relatively low density suggests the possibility of reduced predation pressure on the fish community and may account for the fairly high densities of acanthurids and scarids. The total biomass for the eight families (Fig. 15) provides an assay offish community production in the windward N. A. Sites within the N. A. were relatively productive in fish biomass. A comparison of fish biomass from the same eight families from 60 sites along Jamaica’s north coast (Klomp et al. in press), where decades of intense fishing activity has reduced fish abundance and size, indicates that fish biomass is over seven times greater in the windward islands of the N. A. Although not conclusive from this study, the greater biomass in Saba in part may be a result of reduced fishing pressure associated with the long-term protection within the Saba Marine Park since 1987. 428 The trophic guild classification (Fig. 10) illustrates the functional structure of the fish community as it existed in late 1999. It is important to remember that the forces responsible for shaping this community are a result of past events (e.g., larval recruitment and settlement, predation, environmental conditions) and cannot be identified from a single survey. This assessment does serve, however, as a baseline against which future data can be compared and, collectively with other AGRRA assessments, a point from which current trends in community structure can be detected (see Kramer, this volume). Being able to detect between-site differences in live coral cover within the four geographical areas allows us to partition sites according to their potential for reef development. For example, Acropora has been virtually eliminated from Saban sites SABOl, SABOS, and SAB09 by recent hurricanes (see above). SAB03 is downstream from the only major coastal development on Saba (a man-made harbor and a recently closed rock-crushing facility). Increased sedimentation may be a limiting factor for coral development at SAB03 (Buchan, 1998). The remaining Saban sites offer the greatest potential for reef development and stand to benefit the most by maintaining marine-park protection. Indeed, while the coral assemblages along the west coast are not generally recognized as “true reefs” (Bak, 1977; Nagelkerken, 1981), they have accreted a contiguous, biogenic substratum (e.g., at SAB04, SABOS, and SAB06) and could, arguably, be classified as coral reefs (van’t Hof, 1991). Not surprisingly, the M annularis complex is the dominant coral (41% of all colonies) at these three sites. Our estimates for dominant coral cover (i.e., M. annularis complex = 46% of all colonies) and percent coral cover (mean = 23%) at SAB04 and SABOS, agree with estimates reported by Buchan (1998) for the CARICOMP coral reef site which lies between these two AGRRA sites (CARICOMP; M annularis complex = 43% of all colonies; live coral cover = 24%). Having an area greater than 2,200 km , we did not sample a large enough sector on the Saba Bank to make definitive statements about the condition of its reefs. However, on the basis of our reconnaissance dives in October 1999 and previous reports (Hoetjes et. al., 1999; MacIntyre et. al., 197S; Van Der Land, 1977), it is clear that live coral cover in some sectors exceeds that recorded during our December 1999 assessment and elsewhere approximates our lower estimate («1 1%). An extensive investigation, with the sole purpose of inventorying Saba Bank, would greatly increase our ability to speculate on the potential for reef development in the absence of coastal processes. EUX09 and EUXIO had higher than average values for live coral cover (46%), and were dominated by colonies of Montastraea annularis faveolata having larger than average sizes They may also have a positional advantage being the most southerly of our survey sites in St. Eustatius. Since prevailing coastal currents approach this area from the southeast (ENCAMP, 1980b), stony coral growth at EUX9 and EUXIO may be enhanced by currents introducing relatively nutrient-rich water and/or by receiving less coastal run- off. The two sites surveyed on St. Maarten are routinely exposed to strong currents (SXMOl) and high-energy waves (SXM02), their mean live coral cover was low (1 1%), and they are probably the most threatened reefs we encountered in the windward N. A. Even were hurricanes not a certain hazard, it will be a challenge to afford them enough protection to diffuse the impacts of St. Maarten’s burgeoning coastal development and tourism. 429 ACKNOWLEDGMENTS The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. Special provisions were made by P. Hoetjes and the Netherlands Antilles Government - Department of Public Health and Environment, Winair, Saba Deep Dive Center, Dive Statia, the Saba Tourist Bureau, and Saba Marine Park. We are particularly grateful for field assistance provided by A. Deschamps, K. Marks, C. Moses, S. Steiner, and P. Fletcher. Special thanks for local assistance and logistical arrangements goes to A. Caballero and P. Ellinger and The Nature Foundation St. Maarten, L. Duiveman and St. Eustatius Marine Park, and F. Dilrosun from the N. A. Department of Public Health and Environment. Additionally, appreciation is extended to J. Lang, whose editing and advice has greatly improved this manuscript. REFERENCES Adey, W.H., and R. Burke 1976. Holocene bioherms (algal ridges and bank-barrier reefs) of the eastern Caribbean. Geological Society of America Bulletin 87:95-109. Bak, R.P.M. 1997. Coral reefs and their zonation in Netherlands Antilles. American Association of Petroleum Geologists Bulletin 4:3-16. Bak, R.P.M. , and M.S. Engel 1979. Distribution, abundance and survival of juvenile hermatypic corals (Scleractinia) and the importance of life history strategies in the parent coral community. Marine Biology 54:341-352. Birkeland, C. 1997. Life and death of coral reefs. Chapman and Hall, New York, 536 pp. Buchan, K. 1998. Saba, Netherlands Antilles. Pp. 187-193. In B. Kjerfve (ed.), CARICOMP: Caribbean coral reef seagrass and mangrove sites. UNESCO. Paris. Dilrosun, F.F. 1 999. Saba Bank quarterly progress report on the fish stock assessment project. Report for the Department of Public Health and Environment (VOMIL). Curacao, Netherlands Antilles, 6 pp. ENCAMP 1980a. Saba: preliminary data atlas. Survey of Conservation Priorities in the Lesser Antilles. Eastern Caribbean Natural Area Management Program. 1 8 pp. ENCAMP 1980b. St. Eustasius: preliminary data atlas. Survey of Conservation Priorities in the Lesser Antilles. Eastern Caribbean Natural Area Management Program. 1 8 pp. Gaut, V.C., and J.L. Munro 1983. The biology, ecology and bionomics of the grunts, Pomadasyidae. Pp. 110-141. In: J.L. Munro (ed.), Caribbean Coral Reef Fishery Resources. ICLARM Studies and Reviews 7. Manila, Philippines. 430 Goodson, G. 1976. Fishes of the Atlantic Coast: Canada to Brazil, including The Gulf of Mexico, Florida, Bermuda, The Bahamas, and The Caribbean. Stanford University Press, Stanford, California, 202 pp. Hoetjes, P., M. Vermeij, and A. Caballero 1999. Report on some dives on the Saba Bank. Report for the Department of Public Health and Environment (VOMIT). Curacao, Netherlands Antilles, 3 pp. Website: http://mina.vomil.an/nieuwsbrief.html#sababankreport. Hughes, T.P. 1985. Life histories and population dynamics of early successional corals. Proceedings of the Fifth International Coral Reef Congress 4:101-106. Humann, P. 1992. Reef Creature Identification: Florida Caribbean Bahamas. N. Deloach (ed.). New World Publications, Inc., Jacksonville, Florida, 320 pp. Humann, P. 1993. Reef Coral Identification: Florida Caribbean Bahamas. N. Deloach (ed.). New World Publications, Inc., Jacksonville, Florida, 239 pp. Humann, P. 1994. Reef Fishl Identification: Florida Caribbean Bahamas. 2^^ Ed. N. Deloach (ed.). New World Publications, Inc., Jacksonville, Florida, 396 pp. Klomp, K.D., K. Clark, K. Marks, M. Miller In press. Condition of reef fish on Jamaica’s north coast signals late stages of overexploitation. Proceeding of the Gulf and Caribbean Fisheries Institute 5/^ Annual Meeting. Kojis, B.L., and N.J. Quinn 2001. The importance of regional differences in hard coral recruitment rates for determining the need for coral restoration. Bulletin of Marine Science 69: 967-974. Lewis, S.M. 1980. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Littler, D.S., M.M. Littler, K.E. Bucher, and J.N. Norris 2000. Marine Plants of the Caribbean: A Field Guide from Florida to Brazil. Smithsonian Institution Press, Washington D.C., 263 pp. Littler, M.M., D.S. Littler, and B.E. Lapointe 1 986. 'Baseline studies of herbivory and eutrophication on dominant reef communities of the Looe Key National Marine Sanctuary. NOAA, U.S. Dept, of Commerce, National Ocean Service, Washington, DC. NOAA Technical Memorandum, NOS MEMD 1, 49 pp. MacIntyre, I.G., D.J.J. Kinsman, and R.C. German 1975. Geological reconnaissance survey of Saba Bank, Caribbean Sea. Caribbean Journal of Science 15:11 -20. Meesters, E.H., H. Nijkamp, and L. Bijvoet 1 995. Towards sustainable management of the Saba Bank: a report for the Department of Public Health and Environment (VOMIL). Curacao, Netherlands Antilles, 58 pp. 431 Morrison, D. 1988. Comparing fish and urchin grazing in shallow and deeper coral reef algal communities. Ecology 69:1367-1382. Nagelkerken, W. 1981. Distribution of the groupers and snappers of the Netherlands Antilles. Proceedings of the International Coral Reef Symposium 2:479-484. Nijkamp, H., R. Djohani, and E. Meesters 1 995. The marine environment of St. Maarten. Report by Aid Environment to Dienst VROM, St. Maarten, 24 pp. Oxenford, H.A., and W. Hunte 1990. A survey of marine habitats around Anguilla, with baseline community descriptors for coral reefs and seagrass beds. Report for the Department of Agriculture and Fisheries, Government of Anguilla, 177 pp. Parrish, J.D. 1987. The trophic biology of snappers and groupers. Pp. 405-463. In: J.J. Polovina and S. Ralston (eds.). Tropical snapper and groupers: biology and fisheries management. Westview Press. Boulder, Colorado. Randall, J.E. 1967. Food habits of reef fishes of the West Indies. Studies in Tropical Oceanography 5:665-847. Reeson, P.H. 1983. The biology, ecology and bionomics of the parrotfishes, Scaridae. Pp. 166-177. In: J.E. Munro (ed.), Caribbean Coral Reef Fishery Resources. ICLARM Studies and Reviews 7. Manila, Philippines. Risk, A. 1998. The effects of interactions with reef residents on the settlement and subsequent persistence of ocean surgeonfish, Acanthurus bahianus. Enviromental Biology of Fishes 51 :377-389. Rylaarsdam, K.W. 1993. Life histories and abundance patterns of eolonial corals on Jamaican reefs. Marine Ecology Progress Series 13:249-260. Schmitt, E.F. 1 997. The Influence of Herbivorous Fishes on Coral Reef Communities with Low Sea Urchin Abundance: A Study Among Reef Community Types and Seasons in the Florida Keys. Ph.D. Dissertation, University of Miami. UMI Publishing, Ann Arbor, Michigan. 221 pp. Schmitt, E.F., D.W. Feeley, and K.M.S. Sealey 1998. Surveying Coral Reef Fishes: A Manual for Data Collection, Processing, and Interpretation of Fish Survey Information for the Tropical Northwest Atlantic. Reef Environmental Education Foundation and The Nature Conservancy, Key Largo, Florida, 84 pp. Smith, F.G.W. 1948. Atlantic Reef Corals: A Handbook of the Common Reef and Shallow Water Corals of Bermuda, Florida, The West Indies and Brazil. University of Miami Press, Coral Cables, Florida, 164 pp. 432 Smith A.H., C.S. Rogers, and C. Bouchon. 1997. Status of Western Atlantic coral reefs in the Lesser Antilles. Proceedings of the 8'^ International Coral Reef Symposium 1:351-356. StatSoft, Inc. 1998. STATISTICA for Windows (Computer program manual). Release 5.1. StatSoft, Inc., Tulsa, Oklahoma. Thompson, R., and J.L. Munro 1983a. The biology, ecology and bionomics of the hinds and groupers, Serranidae. In: J.L. Munro (ed.), Caribbean Coral Reef Fishery Resources. ICLARM Studies and Reviews 7. Manila, Philippines, pp. 59-81. Thompson, R., and J.L. Munro 1983b. The biology, ecology and bionomics of the snappers, Lutjanidae. In: J.L. Munro (ed.), Caribbean Coral Reef Fishery Resources. ICLARM Studies and Reviews 7. Manila, Philippines, pp. 94-109. Van der Land, J. 1977. The Saba Bank, a large atoll in the northeastern Caribbean. FAO Fisheries Report No. 200:469-481. van’t Hof, T. 1991. Guide to the Saba Marine Park. Saba Conservation Foundation, N. A., 102 pp. Vaughan, T.W. 1919. Fossil corals from Central America, Cuba, and Puerto Rico, with an account of the American Tertiary, Pleistocene, and Recent coral reefs. Smithsonian Institution U.S. National Museum Bulletin. 103:189-524. Westerman, J.H., and H. Kiel 1961. The geology of Saba and St. Eustatius. Uitgaven Natuurwetenschappelijke Studiekring Voor Suriname en de Nederlandse Antillen, 24. Utrecht, The Netherlands. 175 pp. Williams, D.M. 1991 . Patterns and processes in the distribution of coral reef fishes. Pp. 437-474. In: P.F. Sale (ed.). The Ecology of Fishes on Coral Reefs. Academic Press, Inc. San Diego, California. Wittenberg M., and W. Hunte 1992. Effects of eutrophication and sedimentation on juvenile corals. I. Abundance, mortality and community structure. Marine Biology 1 12:131-138. 433 434 (D X) C 0) T3 c/3 s c/3 cd ^3 C cd (U s •o C3 cd (U cd ■> c/2 C/3 -H « aj 3 a < ' — ^ CO .2 § ■Td N ^ 00 _g o x: ^ e £ = 5 0‘/~>'^^00»r)W^»r)S© CNOfN'^OfNOvd'^’r^ Oir»ooo*/^»r)Oir)ON .'o— '’OOfNO o^oo^oooo^^o-^r^ oooooooor^^ (Nr^rNir^r^cNcNrst^^ \of^oooaN»rj^aN^2 (NOtN'OrTfS’^r^fN — (N(N(N -+C 2 CT' (NO»/-»r--'NOcnI^ B o u ^ ‘I- ■S t: <\j o.s ^ Sou*; ■§ O O •5 c ■£ j c .t: Q O O aj O eg ^ U X H X J CO •C fc o o Oh a < o o 0t^fn \o — fn'©»/^Os‘n(^ms X2X2^!^2i£X2^ 00 c o o. CO S cd o 60 OOli-. o "O c o ^ oa ^ g' K 9J — •S s is 3 ,S o O cn n £ E E o o 5 o o ^ fS '-S ■? S 3 pu > H S S -• -H Os d O 1/3 — d -H os -H (N S W I ^ Is fi^is Hen & Chicks 117 47 ±55 19 ±21 2.0 ±1.5 18 ±27 20 ±30 All St. Maarten 222 37 ± 14 15 ±5 1.3 ±1.2 13 ±7 14 ±8 ‘Primarily Microspathodon chrysurus. 435 CD ■4-> .s (U T3 C3 T3 C on -H C cd (U a s: <3 a a 'O c cC o D l-i 13 o o c o ■*-> on <^-i O cn C (U t3 cn ^ (U on ^ O C on « 'C B cn o T3 cc3 C o ^ 03 OJ ^:z; ■< -o . ^ m ^ (U ^ S £ o •2 2 Q % -C ^ op £ u a oooooooooo so O— iooiriooooo -H-H-H-H-H-H-H-H-H-H oooo'r»oov~»0'^ On OS O-^O(N»r)00s0O mmmmrsrsrsmrs so IT} :i!4^-H4^-H4^-H-f^-H4^ 00 (N — IT) m — (N C/D00COC/DC/300C/3C/5V3 o o o o o 00 _ ^ ^ — V 41 NO 0 0 0 1/1 — ^ — o o -H 4^ 4^ 4 O O O Tf 4i 4^ 4^ 4i -H 4 -H O so m VI m ^ -H 4^ 00 O O -H Tf -i — W1 — V V o o o = vs tn Tf 2 ^ (N cn o o o z 2: 2: CQ CQ CQ o o o o o o o vs vs vs IT) vs vs O o o o o o o o -H 4^ 4i 4^ 4i 4i 4i o o o o o o o 4^ 4^ -H 4^ 4^ 4^ 4^ On ^ 00 Tj- ^ 00 (N cn ^ vs ^ vs -4-* ^ ^ .tS ;?ud:hxjqq£ho< E 5 t « - ■!< s « — < 2 i)CQ (U u 00 00^4-. -a c o E S3 n. P B o , = S >X E ooh22 •< I sa is ,2 V) 55 2 Hen & Chicks SXM02 67 15±21 51±31 34 ±27 2.0± 1.0 33 .18 All St. Maarten 125 8 ± 10 57 ± 8 36 ± 3 1.4 ±1.2 16 ±23 .24 ±,08 ‘Macroalgal index = % relative abundance of macroalgae x macroalgal height 436 Table 5. Mean density and biomass of AGRRA fishes by site in the windward Netherlands Antilles. 437 S O 2 £ 'Sb ^ E (A O c o Q ^ £ ® O 2 5 ©t) ^ B (A O c o Q ^ (A P C3 ^ C O O 2 S 'ob ^ 6 o c o V — • E o S 'Sb ^"e g§ Q S I E E <=> o 2 m 'Sb B §2 Q S c o O 2 m 'ob E 12 Q 5: r-^moovr^so — ON -H rrT}-ON'nro‘r)\OfS'0^ o »r> <^1 o o w-> — i o o o o r4 — ^ 00 ^ On fN — — omNO rs __ sD *— ro 00 r*^ ^ On "Na- rr S: m fs — 00 <^ , ^ ro O r- — (N ^ 00 r- — 41 On On No22NOO-H TT ONr>.^r^oo”2^^^ 'n ov^ov^mw-ioo •— m (N ^ cn rr <^’ Z- -H ^oooogolQo^-H o o o o Z2 OnO'/^ttnOOntTnO-^ S rsooojNO — — O'^ ^ corNifnnNO — NO M On— ■^OOOO— O'— ■" — fNTffO'^r'ifNmtN — o ® >nO M 2 <^■ s :£ :£ 2 cK 2 r-' r-' ^ moomrONOON — f^rfoo NOrTNOONNOOOrrrsONOO -H NO'y^r^r^-'a-mmoor^ — — — fN<_/ ><_/ 'w' ~ t'^ooooc/boococ/bc/bc/boo^ -H CQ ■? VD W i>3 CQ CQ CQ < -H 25 cm maximum diameter) were dominated by Montastraea annularis, Porites astreoides and P. porites. However, live colonies of Acropora palmata, which once flourished in the high- energy shallow-reef zones, had virtually disappeared. The maximum diameter of large reef corals averaged 58 cm which may be indicative of steady juvenile replenishment. The low values of recent partial-colony mortality (<3%) and minor disease occurrences (<6% of colonies) at all sites indicated that the large corals on Horseshoe Reef had experienced no major recent disturbance events. Pale, or partially bleached, colonies on the deeper reefs (about 10% of large corals) were probably still recovering from the 1998 mass bleaching event. Algal communities in the shallower reefs were dominated by crustose coralline algae (>50% relative abundance) whereas macroalgae (mainly Halimeda and Dictyota) were slightly more abundant than crustose corallines at 9-1 1 m. Diadema antillarum was uncommon in the deeper reefs but moderately abundant at 3-4 m. Eighty-one species of fish were recorded at Horseshoe Reef. The assemblage of censused fishes was dominated by herbivorous scarids (parrotfishes) and acanthurids (surgeonfishes). Herbivores, scarids in particular, also accounted for most of the censused fish biomass on the reef. Commercially valuable serranids, lutjanids and haemulids (groupers, snappers, grunts) were present in low densities (<1 individual/ 100m ), indicative of overfishing. INTRODUCTION This study was located in the Tobago Cays, four small, uninhabited islands in St. Vincent-Grenadines (Fig.l) that are partially surrounded by fringing reefs. The Tobago ’Ottawa-Carleton Geoscience Center, Department of Earth Sciences, University of Ottawa, Ottawa, ON, KIN 6N5, Canada. Emails: alice.deschamps@ccrs.nrcan.gc.ca, adesro@uottawa.ca 2 Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI 48824. Eamil: klompkri@msu.edu 440 Cays are protected on their windward sides by Horseshoe Reef, a semicircular, well- developed Holocene bank-barrier reef that is one of the longest (« 4km) in the southern Grenadine Islands (Dey and Smith, 1989). Seaward of its reef flat, the Horseshoe fore reef slopes steeply eastward to depths of about 20 meters. The Grenadines are eroded remnant peaks and ridges of older volcanic islands that were partially or completely drowned during the post-Pleistocene rise in sea level. They arise from the submarine Grenadine Bank which runs NE-SW for about 180 km between St. Vincent and Grenada. The Bank is 15- to 20- km wide and 20-30 meters deep with local shallows at 3-6 m (Dey, 1985). The absence of rivers on these small islands prevents terrigenous sediments from being carried offshore and permits reef growth (Dey, 1985). The Grenadines Bank lies within the Trade Wind belt where strong northeast winds develop during late autumn and winter and easterly winds during spring, summer and early autumn (Clack, 1977). The Equatorial Current moves over the Grenadine Bank, passes through the channels and reefs, and is diverted around the islands giving considerable local variation in current direction and strength (0.3-1. 5 knots) (Dey and Smith, 1989). Tides in the Grenadines have a small range (0.6 m) and are mixed with a semidiurnal component dominating in the Tobago Cays area (Clack, 1977). Their strength, duration and general direction are influenced by local topographic variations present on the sea bottom of the shallow Bank. Temperature and salinity are both relatively constant over Grenadine Bank (Dey, 1985). The Bank is south of the region usually affected by hurricanes. The most violent recent storms to strike the Grenadine islands were Hurricanes Janet in 1955 and Allen in 1980 (Dey and Smith, 1989). No other hurricanes during the past 50 years have seriously affected the Tobago Cays’ reefs (Kurt Cordice, personal communication). The Tobago Cays are 3 km from Mayreau, the closest inhabited Grenadine Island with a population of approximately 250. Union Island, the closest major population center, with approximately 3500, is about 5 km away. The Cays are frequently visited by fishermen harvesting conch and lobster and are now increasingly used by tourists for sailing, snorkeling and diving. Over the past 1 5 years a number of informal reports have indicated that reefs in the Tobago Cays have deteriorated as a result of physical damage from storms, anchors, and fishing gear, as well as from white-band disease, other diseases, and localized nutrient pollution from yachts (Wells, 1988; Smith et al., 1997). The Tobago Cays Marine Park (TCMP) protected area was established in 1998 and is gradually developing its management plans. Horseshoe Reef lies within the TCMP and some fishing regulations have been established. Conversations with local residents revealed, however, that illegal fishing with spearguns is a common practice in the local artisanal fishing community and that any sizable fish, including parrotfish, are targeted. Indeed, we witnessed fishing activity within the marine park on a daily basis during the course of our survey. Except for preliminary descriptions by Lewis (1975), the Tobago Cay reefs have received little scientific attention (Wells, 1988; Smith et al., 1997). Hence, we aimed to characterize the present condition of Horseshoe Reef using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol. 441 METHODS Field data were collected from June 7-16, 1999. Dive sites (Fig.l) that were representative of areas of maximum reef development were selected with the help of true color aerial photographs (scale 1:10 000, March 1991), nautical charts (Hydrographic Office of the United Kingdom, 1999), reconnaissance dives, and the local knowledge of experienced divers and dive-shop operators. Site selection was limited, however, by accessibility to boats and by strong tidal currents. Two shallow reefs (B, D), located between 1 and 5 m in the shallow fore reef, consisted of dead A. palmata pavement, largely encrusted by crustose coralline algae and colonized by Millepora and scattered Porites. Three deeper reefs (A, C, E), located between 8 and 1 5 m on the fore-reef slope, consisted of dead coral pavement, largely colonized various coral species dominated by Montastraea and Porites with scattered patches of Halimeda. Stony corals, algal groups, and Diadema antillarum were censused by one or two divers/survey. The following modifications were made to the AGRRA Version 2.2 protocols (see Appendix One, this volume). The size of individual corals was recorded to the closest 5 cm. Damselfish that were defending territories on censused corals were counted. Sediment in the algal quadrats was lightly brushed away before scoring the abundance of crustose coralline algae. Macroalgal heights were measured to the nearest 0.5 cm. Small (<2 cm diameter) corals of species that are tiny as adults (e.g., Favia fragum) were not counted as “recruits.” Before starting the surveys, pilot transects were conducted in the back reef and results compared to ensure consistency between divers. All fish surveys were made by one diver between 9:00 a.m. and 4:00 p.m. Counts of serranids (groupers) were restricted to species of Epinephelus and Mycteroperca; scarids (parrotfishes) and haemulids (grunts) less than 5 cm in length were not tallied. Roving Diver Transect (RDT) surveys averaged 30 minutes each. Fish biomass was estimated using the length-weight relationships given in Appendix Two (this volume). Field guides included Humann (1994, 1996). RESULTS A total of 60 benthic line transects with 531 corals, 268 algal quadrats, 10 roving diver fish counts and 50 fish-belt transects were conducted on the Horseshoe Reef (Fig. 1, Table 1). Weather conditions were good during most dives with horizontal visibility estimated at approximately 20-25 meters. Stony Corals The assemblage of “large” stony corals (>25 cm maximum diameter) was represented by 16 species and dominated overall by Montastraea annularis (31%), Porites astreoides (23%), P. porites (23%), Montastraea faveolata (5%), Millepora complanata (4%), Colpolphyllia natans (2%) and Siderastrea siderastrea (2%). Nine large coral species predominated at the two shallow reefs (Fig. 2A) with P. astreoides > M. annularis 442 A Shallow (3-4m) n=218 Acropora palmata (9%) Millepora complanata ( 1 0%) Others (1%) Porites porites (21%) B Deeper (9-1 Im) n=313 Others (10% Siderastrea siderea (4% Colpophyllia natans (4% Montastraea faveolata (7%) Porites porites (24%) Porites astreoides (32%) Montastraea annularis (27%) Porites astreoides ( 1 7%) Montastraea annularis (34%) Figure 2. Species composition and mean relative abundance of all stony corals (>25 cm diameter) on (A) shallow, (B) deeper fore reefs at Horseshoe Reef 443 55 50 45 g 40 >> 35 c 30 s O" 25 20 15 10 55 50 45 40 £ 35 w s 30 3 25 u 20 15 10 £ c a> 3 O" 25 cm diameter) of (A) Montastraea annularis, (B) Porites astreoides, (C) Porites porites at Horseshoe Reef. 444 > P. porites > M. complanata ~ standing dead colonies of Acropora palmata (recognizable by their characteristic colony shape, and only present at site B). The deeper reefs, which had a total of 1 5 large coral species, were dominated by M. annularis > P. porites > P. astreoides > M. faveolata > C. natans ~ S. siderea (Fig. 2B). Live stony coral cover averaged about 30% at the shallow reefs and 38% at the deeper reefs (Table 1). The mean diameter of the large colonies among sites ranged from 50 to 65 cm (Table 2). Size-frequency distributions for the three major reef builders (M annularis, P. astreoides, P. porites) were skewed towards smaller colonies (Fig. 3). Larger clumps of P. porites (to 300 cm) and mounds of Montastraea annularis (to 150 cm) were also present but relatively uncommon. Coral recruitment averaged 0.25/0.625 m^ (~4/m^) the shallow reefs and 0.16/0.625 m^ (~2.5/m^) in the deeper reefs (Table 3). Agaricia and Porites were the most commonly seen recruits. Fewer than 4% of the large corals at the shallow and deep sites of Horseshoe Reef showed signs of disease (Table 2). Four percent of the colonies of Montastraea were affected by yellow-blotch disease (YBD) while 2% of all massive corals had black-band disease (BBD). Whereas BBD was only present at one deeper reef (E), YBD occurred at all sites but C. No large corals were completely bleached and the percentage that were either pale or partially bleached varied from about 2.5% to 10.5% in the shallow and deeper reefs, respectively (Table 2). Montastraea accounted for 45% of the bleached corals in shallow reefs and 84% of bleached corals in deeper reefs. However, the overgrowing mat tunicate, Trididemnun solidum, which was present at all five sites, had partially encrusted about 10% (54/531) of the surveyed corals, primarily Porites, Montastraea and, occasionally, Millepora. Partial-colony mortality on the upward-facing surfaces of surveyed corals averaged 2% for recent mortality (hereafter recent mortality) and 25% for old mortality (hereafter old mortality) (Table 2). Frequency distributions (Fig. 4 A, B) reveal that recent mortality for over 95% of these corals was less than 10% at both shallow and deeper sites, whereas fewer than 50% had old mortality values within this range. The most commonly observed sources of recent mortality were predation by Coralliophila abbreviata, parrotfish bites, and incorporation in damselfish algal gardens. Over 60% of the 143 damselfish counted were associated with the M. annularis species complex, with about a quarter of the colonies having at least one resident algal gardener. At the deeper reefs, localized bleaching was associated with dense stands of the calcareous macroalga Halimeda that were observed overgrowing the margins of some stony corals. Algae and Diadema antillarum Algal communities at the two shallow and one of the deeper sites were dominated by crustose coralline algae, whereas macroalgae (particularly the calcareous Halimeda spp., and, to a lesser degree, the fleshy Dictyota spp.) were the predominant algal functional group at the remaining two deeper reefs (Table 3). Average macroalgal canopy height was 1.5 cm (Table 3). Macroalgal indices (macroalgal relative abundance x macroalgal height) were considerably higher in the deeper reefs (~26 at 3-4 m versus ~64 at 9-12 m respectively). Diadema antillarum was present with mean densities of about five individuals/ 100 m^ in the shallower reefs and about 0.5/100 m^ in the deeper reefs. 445 u c a O' V u b o O' O) u b A Shallow (3-4 m) ■ old mortality, n=2 18 □ recent mortality, n=208 L L ■ 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 % old & recent mortality B Deeper (9-11 m) ■ old mortality, n=3 13 □ recent mortality, n=3 12 0-10 10-20 20-30 30-40 40-50 50-60 60-70 70-80 80-90 90-100 % old & recent mortality Figure 4. Frequency distribution for all stony corals (>25 cm diameter) of old partial colony mortality and of recent partial colony mortality at (A) shallow and (B) deeper fore reefs at Horseshoe Reef 446 Fishes A total of 8 1 species of fish were recorded during the 1 0 roving diver surveys («7 hours). The most commonly sighted were planktivorous pomacentrids and herbivorous scarids, pomacentrids and acanthurids, followed by planktivorous labrids (Table 4). Aulostomus maculatus (trumpetfish), present in all surveys, was the most common predator. The fish assemblage quantified in the belt transects was dominated by scarids with 28 and 36 individuals/ 100 m^ having body lengths of at least 5 cm in shallow and deeper sites, respectively (Fig. 5). Acanthurids were present with higher average densities at 3-4 m than at 9-12 m (12 versus 2 individuals/ 100 m^, respectively). Similarly, Microspathodon chrysurus (yellowtail damselfish) were more abundant in shallow reefs (9 individuals/100 m ) than in the deeper reefs (3 individuals/ 100 m ). Commercially valuable species of select serranids, lutjanids (snappers), and haemulids (>5 cm only) were collectively present at densities of less than 1 individual/ 100 m^ (Zn = <20 individuals). It is notable, however, that although lutjanids were less commonly encountered in the belt transects than serranids or haemulids, Lutjanus mahogoni (mahogany snapper) were sighted more frequently (90 percent of dives) than serranids or haemulids during the roving diver surveys (Table 4). rq E o o 2 2 50 40- 30- ■ Shallow sites n=607 Deeper sites n=760 C/5 s: 6 Figure 5. Mean fish abundance (no. individuals/100 ± se) for AGRRA fishes at Horseshoe Reef Other = Microspathodon chrysurus > Bodianthus rufus. 447 A Carnivores s it 3 C 40 - 30 - 20 - 10 Shallow sites (n=8) Deeper sites (n=10) ^Dotted line indicates proportion of predators composed of Serranidae and Lutjanidae (below line) vs. Haemulidae (above line). No Haemulidae >40 cm were sighted. 0-5 6-10 11-20 21-30 31-40 >40 Size (cm) B Herbivores Size (cm) Figure 6. Size frequency distribution of (A) carnivores (lutjanids, select serranids, haemulids >5 cm), (B) herbivores (all acanthurids, scarids >5 cm, Microspathodon chrysurus) at Horseshoe Reef 448 Although commercially significant predators were rare in our belt transects, half (shallow) to most (deeper) of the groupers and snappers encountered ranged in size from 21-30 cm (Fig. 6A). Serranids (mean = 29 cm) were about twice as long as lutjanids (mean = 1 6 cm), which overlapped in size with the haemulids (all of which were >5 cm) (Fig. 7A). The lengths of half or more of each of the herbivorous groups (scarids >5 cm, acanthurids and Microspathodon chrysurus) were in the 1 1 -20 cm length class and averaged about 15 cm (Figs. 6B, Fig. 7B). Of the AGRRA fishes surveyed in the belt transects, the biomass of the herbivorous guild was 90 percent greater than that of the commercially valuable carnivores (Fig. 8). The scarid biomass was evenly distributed between shallow and deeper sites. Biomass estimates of acanthurids and M. chrysurus were considerably greater in the shallow reefs than in the deeper fore reefs (Table 5). DISCUSSION The coral assemblage and coral cover on Horseshoe Reef are typical of high- energy, windward Caribbean reefs and, except for the near disappearance of live A. palmata from the reef flat and shallow fore reef, very similar to the descriptions of Lewis (1975) (some juvenile A. palmata are present in the back-reef area immediately behind the reef crest). The presence of standing dead colonies of A. palmata in the shallow fore reefs is a good indication that they were not killed by hurricanes. Hence, we concur that their demise at Horseshoe Reef, as elsewhere in St. Vincent-Grenadines, was probably due to white-band disease (Antonius, 1981; Aronson and Precht, 1997). The positively skewed size distribution of the three major reef-building corals may be an indication of a reef system with adequate juvenile input (Bak and Meesters, 1998) since large-scale fragmentation by hurricanes seems unlikely on Horseshoe Reef (see Introduction). The slightly lower density of coral recruits on the deeper reefs is probably related to the generally higher abundance of macroalgae as well as the relatively lower abundance of crustose coralline algae and the near absence of Diadema antillarum (Birkeland 1977; Pearson, 1981; Hughes et al., 1987). None of the factors contributing to recent mortality are responsible for major damage to the corals at Horseshoe Reef. The pale or partially bleached corals seen during our survey, particularly on the deeper reefs, were possibly still recovering from the 1998 Caribbean- wide bleaching event (Strong et al., 1998) that affected reefs in the Tobago Cays (Kurt Cordice, personal communication). Halimeda, engaged in marginal overgrowth of some stony corals, may also have contributed to the higher occurrence of bleaching on these deeper reefs. Competitive overgrowth by organisms such as Trididemnun solidum, which was not noted in the Tobago Cays by Lewis (1975), may be influenced by environmental perturbations that reduce the efficiency of affected organisms to defend themselves. Thus, its relatively high abundance (on 1 0% of the large colonies) at Horseshoe Reef is cause for concern. Although the effects of T. solidum were scored as “old mortality” in accordance with the AGRRA protocol, there were many cases for which we suspected that loss of coral tissues had occurred within recent months. Had T. solidum been recorded under “recent” mortality, our estimates of the latter would have been substantially larger. 449 A Carnivores WD e 5 cm) Lutjanidae Haemulidae (>5 cm) Serranidae (Epinephelus, Mycteroperca) Herbivores Commercially Valuable Predators n = 2230 n= 18 Figure 8. Mean biomass of key herbivores and carnivores at Horseshoe Reef. □ Horseshoe Reef □ Windward N.A. Figure 9. Mean total length (± se) of common scarids at Horseshoe Reef and from the windward Netherland Antilles (Kloomp and Kooistra, this volume), n = number of sites; number of individual fishes is in parentheses. 451 Algal community structure is greatly affected by the spectrum of grazers present on an individual reef and by fishing pressures that significantly alter these natural grazing patterns (McClanahan and Muthiga, 1 998). Macroalgae can easily overgrow smaller algal turfs and crustose coralline algae (Lewis, 1986). We suspect that the herbivores (e.g., scarids, D. antillarum; to a lesser degree, acanthurids and M chrysurus) on the shallow fore reefs at Horseshoe Reef are maintaining a benthic algal assemblage in which crustose corallines are spatial dominants. The somewhat greater proportion of turf algae at 3-4 m relative to 9-12 m may be a result of higher productivity rates which have been attributed to greater photosynthic activity in shallower habitats, to the forces of herbivory that enhance growth by maintaining algal assemblages in an early successional stage, and to recycling of nutrients by herbivores back into the benthic community (Hatcher, 1997). Grazing intensity generally decreases with depth on fore reefs (Morrison, 1988) which may explain the relatively higher densities and biomass (Table 5) of herbivorous acanthurids and M. chrysurus at 3-4 m than at 9-1 1 m. The macroalgae that predominated at two-thirds of the deeper reefs {Halimeda, Dictyota), where macroalgal grazers (particularly Diademd) are relatively less abundant, are of genera known to be avoided by herbivorous fishes (Schmitt, 1998). A reduction in herbivorous grazing pressure allows macroalgae to outcompete sessile reef invertebrates for space and can ultimately result in the degradation of coral communities (Hughes, 1989; Hughes et al., 1987). However, live stony coral cover was higher on the deeper fore reefs than in the shallow at the time of our survey. As reported in Table 4, L. mahogoni was one of the 25 most frequently sighted species in the roving diver surveys, appearing in 9/10 surveys, yet was not seen in any of the belt transects. Its presence in the roving diver surveys was recorded as “few” (i.e., 2-10 individuals) in seven of these dives and as “many” (i.e., 11-100 individuals) in two dives. This may be an indication that the belt transect method is not adequately quantifying fish such as L. mahogoni, which are inherently wary of divers, or that a larger sample size is needed to adequately quantify species that are present but not necessarily abundant. Of additional interest from the roving diver surveys, Chromis multilineata (brown chromis) and C. cyanea (blue chromis) were recorded as being the most frequently sighted and “abundanf’ (>100 individuals) fishes on Horseshoe Reef The prevalence of these planktivores may be an indication of reduced predation in the presence of an adequate food source. However, they are commonly found throughout the greater Caribbean strategically feeding above reefs where plankton is plentiful (Bohlke and Chaplin, 1993). The Scaridae was the dominant fish family present in our belt transects on Horseshoe Reef (Fig. 5, Table 5). Scarids were represented primarily by Sparisoma viride (stoplight parrotfish) with lesser densities of S. aurofrenatum (redband parrotfish), Scarus taeniopterus (princess parrotfish), S. vetula (queen parrotfish) and S. croicensis (striped parrotfish) (Table 6). Average total lengths for S. croicensis and S. aurofrenatum were similar at Horsehoe Reef and in the windward Netherlands Antilles (N.A.) which have received a lesser degree of fishing pressure (Klomp and Kooistra, this volume). S. viride and S. taeniopterus, however, were both significantly smaller on average (F-test; p=0.001 and p=0.044, respectively) than their conspecifics in the windward NA (Fig. 9). Although we cannot conclusively determine from this assessment the cause for the relatively small size of these parrotfishes in Horseshoe Reef, it may be further evidence of high harvesting levels. 452 Reduced abundance or biomass and a decreased size structure in targeted fishes as a result of intense fishing has been well-documented (Bohnsack, 1982; Munro, 1983). It was not common to see fish larger than 30 cm total length on Horseshoe Reef (Fig. 6) and we suspect that this is an effect of larger fish being targeted for harvest. Additionally, overfishing likely explains the low densities of grouper, snapper and grunts on Horseshoe Reef since these are favored food fishes. It is possible that these common carnivores have been replaced by other species, such as the “voracious predator” Aulostomus maculatus (trumpetfish), which is known to feed on young acanthurids and haemulids (Randall, 1967). A. maculatus are common residents of Caribbean-area reefs but ordinarily are not abundant (Bohlke and Chaplin, 1993). Randall (1967) found trumpetfish in the stomachs of snappers and groupers. Given the paucity of serranids and lutjanids, we suspect that the Horseshoe Reef population of^. maculatus is thriving due to reduced predation pressure and adequate prey supplies. This phenomenon may warrant further investigation if marine park managers seek to restore populations of snappers, groupers or grunts. The Vincentians rely on their artisanal fishery as a source of dietary protein. A restored population of commercially valuable carnivores within the marine park could serve as a source stock of fish and possibly create a spillover effect into legal fishing areas, providing harvestable resources for the fishing community. In sum, strengthening and enforcing fishing regulations in the Tobago Cays Marine Park are critical steps to avoid further degradation of its fish assemblages. The AGRRA survey, being the first quantitative, assessment of Horseshoe Reef, provides baseline information against which to measure future changes in its condition. ACKNOWLEDGMENTS We would like to express our sincere gratitude to Philip and Tricia Kramer for their help in training and as well as their guidance with the AGRRA protocols. We would also like to acknowledge Judy Lang for her comments and assistance with the manuscript. Many thanks to the Grenadine Dive staff and the Tobago Cays Marine Park staff for their assistance during field operations. The senior author received an Ontario Graduate Scholarship. Field work was partially supported by a National Sciences and Engineering Research Council of Canada grant to Andre Desrochers. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. REFERENCES Antonius, A. 1981 . Coral reef pathology: A review. Proceedings of the Fourth International Coral Reef Symposium 2:3-6. Aronson, R.B., and W.F. Precht Stasis, biological disturbance, and community structure of a Holocene coral reef Paleobiology 23:336-46. 453 Bak, R.P.M., and E.H. Meesters 1998. Coral population structure: the hidden information of colony size frequency distribution. Marine Ecology Progress Series 162:301-306. Birkeland, C. 1977. The importance of rate of bio-accumulation in early successional stages of benthic communities to the survival of coral recruits. Proceedings of the Third International Coral Reef Symposium 1:15-21. Bohlke, J.E., and C.G.G. Chaplin 1993. Fishes of the Bahamas and Adjacent Tropical Waters. 2"^ edition. University of Texas Press, Austin, Texas. 771 pp. Bohnsck, J.A. 1982. Effects of piscivorous predator removal on coral reef fish community structure. Pp. 258-267. In: G.M. Cailliet and C.A. Simenstad (eds.), Gutshop ’81. Fish food habits. Proceedings of the Third Pacific Workshop, 6-9 December 1981, Pacific Grove, CA. Washington Sea Grant Publication, University of Washington, Seattle. Clack, W.J.F. 1977. Recent carbonate reef sedimentation off the coast of Carriacou, West Indies. Ph.D. thesis, McGill University, Montreal, 218pp. Dey, S. 1985. Compositional and textural analysis of beach and shallow-marine carbonate environments in the southern Grenadines, Lesser Antilles volcanic island arc: Facies model for ancient island-arc limestones. M.Sc. thesis. Queen’s University, Kingston, Ontario, 281pp. Dey, S., and L. Smith 1 989. Carbonate and volcanic sediment distribution patterns on the Grenadines band. Lesser Antilles island Arc, East Caribbean. Bulletin of Canadian Petroleum Geology 37:18-30. Hatcher, B.G. 1997. Organic production and decomposition. Pp. 140-174. In: C. Birkeland (ed.). Life and Death of Coral Reefs. Chapman and Hall, New York. Hughes, T.P. 1989. Community structure and diversity of coral reefs: the role of history. Ecology 70:275-279. Hughes, T.P., D.C. Reed, and M.J. Boyle 1987. Herbivory on coral reefs: community structure following mass mortality of sea urchins. Journal of Experimental Marine Biology and Ecology 1 13:39-59. Humann, P. 1994. Reef fish identification: Florida, Caribbean, Bahamas, 2"^ Edition. New World Publications, Inc., Jacksonville, FL, 400 pp. Humann, P. 1996. Reef coral identification: Florida, Caribbean, Bahamas. 2"^^ Edition. New World Publications, Inc., Jacksonville, FL, 239 pp. Hydrographic Office of the United Kingdom 1999. Nautical charts of the Middle Grenadines Canouan to Carriacou, Windward Islands. Imray Laurie Norie & Wilson Ltd., England. 454 Lewis, J.B. 1975. Preliminary description of the coral reefs in Tobago Cays, Grenadines, West Indies. Atoll Research Bulletin 17B:2-14. Lewis, S. 1986. The role of herbivorous fishes in the organisation of a Caribbean reef community. Ecological Monographs 56:183-200. McClanahan, T.R., and N.A. Muthiga 1998. An ecological shift in a remote coral atoll of Belize over 25 years. Environmental Conservation 25:122-130. Morrison, D. 1988. Comparing fish and urchin grazing in shallow and deeper coral reef algal communities. Ecology 69:1367-1382. Munro, J.L. 1 983. Caribbean coral reef fishery resources. ICL ARM Studies and Reviews 7, Manila, Philippines. 276 pp. Randall, J.E. 1967. Food habits of reef fishes in the West Indies. Studies in Tropical Oceanography 5:655-847. Strong, A.E., T.J. Goreau, and R.L. Hayes 1998. Ocean Hot Spots and coral reef bleaching: January-July 1998. Reef Encounters 24:20-22. Smith , A.H., C.S. Rogers, and C. Bouchon 1997. Status of Western Atlantic coral reefs in the Lesser Antilles. Proceedings of the International Coral Reef Symposium 1:351-356. Wells, S. 1988. St. Vincent. Pp. 289-294. In: S. Wells (ed.). Coral reefs of the world: Atlantic and Eastern Pacific United Nations Environment Program and International Union for Conservation of Nature and Natural Resources. Table 1. Site information for AGRRA stony coral, algal and fish surveys on Horseshoe Reef, Tobago Cays, St. Vincent. 455 «+-l "O c o a Urn o o O 03 dj m (N 2.5 in 6.5 o dJ >. c o s 00 X) TD cd dj D T? 0) OX) IT) n os O o o c3 03 c 03 s c/x o (N Al o O O in in r-- t/5 * -n in (N -i cd "O 13 j . in O if) in in in "5 c On (U Oii o o o o in o c/5 1— ro n "S -H in ± o ± o o o o SP O O s 5 cm) M icrospathodon chrysurus Haemulidae (>5 cm) Lutjanidae Serranidae All Shallow 1081 ±364 3827 ±487 1145 ±354 84 ±228 6±7 22 ± 12 All Deep 269 ± 67 4217 ±605 330±231 28 ± 128 4±6 198 ±25 Table 6. Mean density of the >5 cm scarids, by site on Horseshoe Reef. Site Density (# individuals/100 m^) S. aurofrenatum (redband) (n=177) S. taeniopterus (princess) (n = 140) S. viride (stoplight) (n = 323) S. croicensis (striped) (n= 112) S. vetula (queen) (n= 156) Shallow B 4.3 1.3 11.5 5.3 1.8 D 6.7 2.7 14.0 2.7 5.0 Deep A 5.3 3.0 5.3 2.3 1.2 C 8.7 8.0 12.0 5.5 17.0 E 4.5 8.3 11.0 2.8 1.0 All sites mean ± se 6.2 ± 2.2 6.4 ± 3.0 9.4 ± 3.6 3.6 ± 1.7 6.4 ± 9.2 459 Plate llA. Reef condition is strongly dependent on the interplay of complex relationships involving stony corals, herbivores and benthic algae. Algae are assessed in the AGRRA benthos protocol as the relative abundance of several key functional algal groups, including macroalgae like the Dictyota shown here, in relation to coral condition and herbivorous fishes. (Photo Robert W. Steneck) Plate IIB. Crustose coralline algae that grow around and between coral fragments may eventually immobilize loose pieces of coral rubble and may serve as recruitment sites for coral larvae. Recruitment is estimated as the number of small (<2 cm diameter) stony corals visible in the algal quadrats, like those shown here. (Photo Robert W. Steneck) 460 from R.N. Ginsburg in P.A. Scholle, and N.P. James (1995). ASSESSMENT OF THE CORAL REEFS OF THE TURKS AND CAICOS ISLANDS (PART 1: STONY CORALS AND ALGAE) BY BERNHARD RIEGL/ CARRIE MANFRINO,^ CASEY HERMOYIAN,^ MARILYN BRANDT,^ and KAHO HOSHINO^ ABSTRACT Major constituents of the benthic reef community (stony corals, algae) were assessed in 28 reefs on the Caicos, Turks and Mouchoir Banks. Living stony coral cover ranged from 8-28%, averaging 18% overall. Montastraea annularis and M. faveolata of “intermediate” sizes (<100 cm) dominated all examined reefs. Live Acropora palmata and A. cervicornis were scarce. The most frequently recruiting scleractinians were Porites astreoides and Agaricia agaricites; Montastraea recruits were uncommon. Old partial-colony mortality (overall mean=23%) was greater than recent partial-colony mortality (mean=3%). Crustose coralline algae and turf algae were generally more abundant than macroalgae. Mouchoir Bank, with the most isolated reefs, was in relatively poor condition, which suggests that remoteness alone does not protect coral reefs. ' National Coral Reef Institute, Oceanographic Center, Nova Southeastern University, 8000 N. Ocean Drive, Dania FL 33004 and Institute for Geology and Paleontology, Karl-Franzens-University Graz, 8020 Graz, Austria. Email: rieglb@nova.edu 'y Department of Geology and Meteorology, Kean University, 1000 Morris Ave., Union, N.J. 07083. ^University of Michigan, Department of Geological Sciences, Ann Arbor, MI 48109. ^ NCORE, Rosenstiel School of Marine and Atmospheric Science, 4600 Rickenbacker Causeway, Miami, FL 33149-1098. ^ Formerly Bren School of Environmental Science and Management, University of California, Santa Barbara, CA 93106-5131. 462 INTRODUCTION The Turks and Caicos Islands (TCI), which lie at 21° to 22° N and 71° to 72° 30' W, consist of 8 islands (7 of which are inhabited) and approximately 40 low-lying cays distributed among two banks (Turks Bank, Caicos Bank) plus part of the entirely submerged Mouchoir Bank (Fig. 1). Over 300 km of coral reef surround the Turks and Caicos Islands (Wells, 1988). The prevailing easterly trade winds (see windrose for the TCI area in Fig. 1) create a clear differentiation on the banks into a windward eastern side with generally choppy conditions and a leeward western side that is usually calm. The banks have narrow, discontinuous, shelf-edge reef (SER) systems (sensu Blanchon and Jones, 1997) of variable depth, relief, and stony coral abundance (Chiappone et al., 1996). Along the western parts of the Caicos and Turks Banks, shallow fringing reefs are developed shoreward of the SERs. Shallow patch reefs also surround many of the islands and cays. Underwater visibility is considered good everywhere. The reefs and banks of the Turks and Caicos Islands have been studied by Wanless and Dravis (1989), Sullivan et al. (1994), Gaudian (1995, unpublished report), Chiappone et al. (1996) and Steiner (1999). In the context of the ongoing general deterioration of reef health in the entire Caribbean basin (e.g., Ginsburg, 1994), these isolated islands, with relatively small human population pressures, are of particular interest as landmark study sites. Information to date indicates that the Turks and Caicos reefs are generally in good condition with some pollution impacts evident near the islands of Providenciales and Grand Turk (e.g., Sullivan et al., 1994; Lang et al., 1998; Steiner, 1999; Woodley et al., 2000). Hence, they can be used for comparison with other sites subjected either to direct continental influences or to higher impacts, both natural and anthropogenic. This study presents; 1) the August 1999 Atlantic and Gulf Rapid Reef Assessment (AGRRA) results for benthic reef condition; 2) an evaluation of differences between shallow versus deep reefs and between windward versus leeward reefs; and 3) a qualitative comparison of reefs on the three banks (which are known to experience different levels of resource extraction). Our AGRRA fish surveys for the Turks and Caicos Islands are presented by Hoshino et al. (this volume). METHODS Survey sites (Fig. 1) were selected with the assistance of locally available diving and sailing maps, charts (British Admiralty, U.S. Navy), maps in publications, and aerial photographs. We chose strategically accessible reefs (e.g., at established dive sites with mooring buoys) that were considered representative of special interests (i.e., reported to be heavily impacted, or of touristic, fisheries and/or conservation value). Although an effort was made to space sites as evenly as possible within all available exposures and reef types. 463 small boats and prevailing sea conditions restricted our surveys to areas of moderate exposure and/or short traveling distance. However, a mix of 1 1 moderately exposed and 17 sheltered reefs were obtained. The northern side of the Caicos Bank and much of Mouchoir Bank were not exhaustively investigated. On Turks Bank we sampled all available habitats within the appropriate depth intervals, but largely ignored the southern area south of Salt Cay. Since an Acropora palmata reef-crest zone was not encountered in any of the areas examined, we surveyed three shallow patch reefs at depths of 2. 5-6. 5 m (Table 1). The patch reefs had been constructed primarily of A. palmata, still had some live colonies of this species, and were considered representative of several other patch reefs that we also visited. Elsewhere we made qualitative notes of the abundance of A. palmata. The remaining surveys were located in depths of 9.5-22.5 m on the seaward margin of spurs in the SERs. Nine were high-relief (>5 m) and 16 had lower relief (<5 m) but all showed groove-and-spur morphology with sand-filled channels running between the reef lobes. Three divers executed the AGRRA Version 2.1 benthos protocol (see Appendix One, this volume) using the following modifications: pockets of sand underlying the transect line were not measured; assessments were made for each stony coral of 10 cm or greater diameter beneath the transect line; colony height and diameter were measured to the nearest 5 cm or, when possible, the nearest cm. Pontes furcata and P. divaricata were not separated from P. porites, and species of Agaricia were not determined but about 90% of the surveyed corals are thought to have been yi. agaricites with most of the rest consisting primarily of A.fragilis and A. humilis. Diseases were characterized by criteria established by Antonius (1995), Santavy and Peters (1997), and Peters (1997). We looked for damselfish tending algal gardens on the individually surveyed corals but none were recorded. Species that are small as adults (e.g., Favia fragum) were not included in the counts of stony coral “recruits.” Sediment was removed from the algal quadrats by fanning the substratum two or three times by hand after scoring the cover of algal turfs and macroalgae but prior to estimating the abundance of crustose coralline algae. Absolute algal abundance estimates frequently exceeded 100% since each layer was estimated separately. Numerous consistency checks were performed. Prior to beginning the survey, all divers performed measurements on the same transect and the results were compared. This process was repeated until results were homogeneous within the group. In total, five training transects were necessary. During the surveys, divers repeatedly discussed coral identification and interpretation of mortality, disease, algal cover, etc. The field guide used for identification of marine organisms was Human (1993). For statistical evaluations, the reefs were grouped into three ecological units: shallow Acropora palmata-hnWX patch reefs; high-relief SERs; and low-relief SERs. All data were found to be normally distributed with Kolmogoroff-Smimow one-sample tests for normality of distribution. Parametric testing statistics were used to compare groups by means of the student’s t-test or one-way analysis of variance and Tukey's post-hoc test to 464 identify significant groupings. Differences between the three banks were not tested for significance since the sample size (n=2) on Mouchoir Bank was not representative. The SERs were also grouped for testing according to expected exposure regime with all reefs on the eastern sides of the banks considered “moderately exposed ” and those on the western side to be “sheltered.” The three patch reefs all were in exposed locations, hence “windward” versus “leeward” comparisons were not possible for this habitat type. RESULTS Stony Corals A total of 3,270 corals were surveyed in 289 transects on three banks and around 1 1 islands in the Turks and Caicos Islands. Live stony coral cover averaged ~18% overall (Table 1). No significant differences were found between the patch reef and the high-relief or low-relief SERs (ANOVA, F=1.749, p=0.195), despite clear evidence of previous A. palmata mortality (numerous large skeletons in the framework) in the former. Differences of exposure were just significant (t-test, F=4.3, p=0.05), the percentage of live coral cover being higher in the moderately exposed SERs (mean=20.5, sd=6.8, n=8) than in the sheltered SERs (mean=14.6, sd=3.3, n=17). The very low coverage (7.5%) seen on one high-relief SER (TC5) was largely due to a local limitation in the amount of suitable habitat as the spurs were dissected into patches each <10 m across and separated from the others by pockets of sand. Scleractinian growth, particularly on Turks Bank, was most profuse in the area immediately adjacent to the platform margin. In some instances two platform edges were found, a shallower rim at 10-15 m depth and a deeper edge seaward of a small (10-50 m wide) plateau at around 30-35 m (best developed near South Caicos). Stony coral growth was always densest on the outer edge of the shallower shelf Steep slopes exhibited few scleractinians but in many areas dense populations of black corals, Cirrhipathes sp. and Antipathes spp., were observed (particularly in TC 7 on Turks Bank). On gentler (<40° inclination) reef slopes, stony corals were abundant to depths of 50 m on Turks Bank (in TC8), the platy coral facies in places (e.g., TCI 9) being well over 50% at 25-30m depth. Of the stony corals that were >10 cm in diameter, A. palmata was most common in the patch reefs, while poritids {Porites pontes, P. astreoides) were more abundant here than in the deeper SERs (Fig. 2). Indeed, P. astreoides became increasingly common with decreasing water depth (and thus increasing hydrodynamic exposure). Montastraea annularis and M. faveolata were the most abundant stony corals in deeper water with the M. annularis complex constituting about 40% of all the colonies in the SERs (Fig. 2). A. cervicornis was present but rare (<1% in SERs) on the Caicos and Turks Banks. It is interesting to note that Dendrogyra cylindrus, which is generally uncommon in most 465 A. palmata patches (N=305) High relief shelf edges (N=l 1 1) Low relief shelf edges (N=1860) AC=A. cervicornis, AG=Agaricia spp., AP=Acropora palmata, C\A=Colpophyllia natans, Y)EHT)KO=Dendrogyra cylindrus, 'DL=Diploria labyrinthiformis, DS=D. sthgosa, 'E\}'$)=Eusmilia fastigiata, F¥=Favia fragum, lSO=Isophyllastrea sp., M.A=Montastraea annularis, MFA=M faveolata, MC=M cavernosa, MFR=M franksi, MAD=Madracis spp., MEA=Meandrina meandrites, y[\L=Millepora spp., MY C=Mycetophyllia ferox, FA=Porites astreoides, PP=P. porites, SS=Siderastrea siderea, F>YE?W=Stephanocoenia intersepta Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>10 cm diameter) in Acropora palmata patch reefs, low-relief shelf-edge reefs and high-relief shelf-edge reefs in the Turks and Caicos Islands. Caribbean areas, obtained counts of 1-2% overall with no apparent preference for any particular depth zone and was particularly conspicuous on Turks Bank. For colonies >10 cm in diameter, the average maximum diameter (Table 2) ranged between 26 cm in a high-relief SER (TC5) and 103 cm in a patch reef (TC9); their average maximum height varied from 16.5 cm (in TCI, a high-relief SER) to 66 cm (in TC9). Average maximum diameter and height were both significantly higher in the patch reefs than in the deeper reefs (ANOVA for greatest diameter F=18.5, p<0.001; for greatest height F=14.7, p<0.001), but no significant differences in size were found between the high-relief and low-relief SERs. Nor were any differences in diameter or height found between moderately exposed and sheltered SERs (t-tests, f=0.71, p=0.401 for diameter; F=1.91, p=0.181 for height). Amongst the more common corals, Acropora palmata showed a polymodal size distribution which was somewhat skewed towards the larger (>100 cm) size classes (Fig. 3A,B). The size-frequency distributions of Montastraea annularis and M. faveolata were 466 70. 60. 50. 40. .10. 20 10 0 1 I li I MiNiliiiti iii.’-ii liiV L^iiiiJta (N=i|>tK) □ Itigh Tciici l^l' R. □ l-0 OO la 0 n H p '■i t % i 1 ft 1 1 . Ai 1 j d *4Mll C't Mi.iiilaslf acii aitnulinis (M-TOVt Q High rcliei SFR Q l-nw relrclSFR ■ A. palmdu pslcbi»i ,rO ,ri|]l,.o -<2. c N" syt‘ class (chunKHo’. cm | i:u UHli to -10 :o oil vSi DjAfaricui si''p. (N=3ISt D High relic! SLR e i.10 cm diameter colonies of (A) Acropora palmata, (B) Montastraea faveolata, (C) M annularis, and (D) Agaricia spp. in Acropora palmata patch reefs, low-relief shelf-edge reefs and high-relief shelf-edge reefs in the Turks and Caicos Islands. 467 Figure 3B. Size-frequency distributions of >10 cm diameter colonies of (A) Acropora palmata, (B) Montastraea faveolata, (C) M annularis, and (D) Agaricia spp. pooled for all reefs in the Turks and Caicos Islands. 468 clearly skewed to the “intermediate sizes” (20-70 cm for M annularis, 20-90 cm for M. faveolata). Most of the Agaricia spp. (primarily^, agaricites) were less than 30 cm in maximum diameter. The eolonies of A. cervicornis in the Caicos and Turks Banks were relatively small (rarely >1 m in diameter). The density of stony coral recruits (Table 3) in the 1,156 algal quadrats ranged from 0.02/0.0625 m^ in a patch reef (TC9) to 0.8/0.0625 m^ in a low-relief SER (TC27). By far the most common (Fig. 4) were Porites astreoides followed by Agaricia spp. (mostly yi. agaricites). Montastraea annularis and M. faveolata were present but in low abundance. No acroporid recruits were encountered during the surveys. Recruit density did not differ significantly between the patch reefs and either the high-relief or the low-relief SERs (ANOVA, F=2.92, p=0.072). There were no differences in recruitment between the moderately exposed and sheltered SERs (t-test, F=0.22, p=0.641). Coral Recruits A. palmata patches (N=23) Low relief shelf edges (N=205) pp 8% PA 37% MYC 1% AC=/1. cervicornis, AG=Agaricia spp., A?=Acropora palmata, CH=Colpophyllia natans, DENDRO=£)e«(irag>’ra cylindrus, DL=Diploria labyrinthiformis, DS=D. strigosa, E\3?>=Eusmilia fastigiata, FF=Favia fragum, lSO=lsophyllastrea sp., MA=Montastraea annularis, MFA=M faveolata, MC=M cavernosa, MFR=M franksi, ]s/[AF)=Madracis spp., MEA=Meandrina meandrites, MlL=Millepora spp., MYC=Mycetophyllia ferox, FA=Porites astreoides, ??=/*. porites, SS=Siderastrea siderea, SYFFW^Stephanocoenia intersepta Figure 4. Species composition and mean relative abundance of all stony coral recruits (S2 cm diameter, excluding species that are small as adults) in Acropora palmata patch reefs, low-relief shelf-edge reefs and high-relief shelf-edge reefs in the Turks and Caicos Islands. High relief shelf edges (N=l 13) STSPH 469 Stony Coral Condition On average, nearly 5% of the >10 cm diameter stony corals on each reef were diseased. The percentages of diseased stony corals were highest in one site on Mouchoir Bank (17% in TC 12) and in two sites on Caicos Bank (13% in each of TC26 and TC27), while the lowest values were found on Turks Bank (Table 2). The percent of affected colonies did not differ significantly between the patch reefs and either type of SER (ANOVA, F=0.223, p=0.802). Moreover, there were no differences between the moderately exposed and sheltered SERs (t-test, F=0.05, p=0.998). Healthy patches of Acropora palmata andH. cervicornis were only encountered in Caicos Bank on the southwestern (between Providenciales and West Caicos) and southeastern (at TCI 5) sides, and in the two patch reefs on the eastern side of Turks Bank (TC9, TCIO). Three of the small colonies of A. cervicornis in the SER reefs (one each in TC3, TC26 and TC28) had white-band disease, and three of the patch-reef A. palmata (two in TC9, one in TCIO) exhibited similar characteristics. Only a few cases of black-band disease were encountered. White plague was common, and about 75% of the 130 diseased colonies belonged the Montastraea annularis complex (M faveolata 43%, M. annularis 27%, M. franksi 5%). No bleaching at all was captured in the dataset; neither was any damage by damselfish observed. Mortality patterns (as a percent of affected upper colony surfaces) were somewhat different between shallow and deeper water (Fig. 5). Values for recent partial-colony mortality (hereafter recent mortality) varied from <0.5% in a patch reef (TCIO) to 7.5% in a low-relief SER (TCI 8) and values of old partial-colony mortality (hereafter old mortality) from 12.5% in a high-relief SER (TC2) to 47% in a patch reef (TC9). Percentages of both old mortality and total (recent + old) mortality were slightly higher in the moderately exposed SERs (old mortality mean=24.9, sd=7.9; total mortality=36.1, sd=3.9) than in the sheltered SERs (old mortality=20.4, sd=3.4; total mortality=31.3, sd=3.9). Recent mortality showed no significant differences between the patch reefs and the SERs (ANOVA, F=1.045, p=0.367). Both old mortality and total mortality differed significantly between the patch reefs and the high- and low-relief SERs, although the latter did not differ from each other (ANOVA for old mortality F=7.33, p=0.03; for total mortality F==10.2, p<0.001). Recent mortality did not differ between the moderately exposed and sheltered SERs (t-test, F=4.05, p=0.056) whereas significant differences were found in old mortality (t-test, F=10.4, p=0.004) and total mortality (t-test, F=8.5, p=0.008), being higher in the sheltered sites than in the moderately exposed SERs. No examples of stony corals having experienced 1 00% recent mortality were encountered. Less than 5% were “standing dead” (100% mortality and still in original growth position) (Table 2), except in two of the patch reefs (TC9-20.5%, TClO-15.5%) where much of the reef framework was made of large, long-dead skeletons of Acropora 470 Figure 5. Log-frequency distributions of (A) recent partial colony mortality and (B) old partial colony mortality of all stony corals (>10 cm diameter) in Acropora palmata patch reefs, low-relief shelf-edge reefs and high-relief shelf-edge reefs in the Turks and Caicos Islands. palmata. The differences in standing dead corals between the patch reefs and the SERs were significant (ANOVA, F=21.77, p<0.001). On the Mouchoir Bank, only isolated and badly damaged ridges of A. palmata were observed in shallow habitats and in many cases the coral skeletons were heaps of large rubble. Diseases (possibly including aspergillosis) in sea fans and other gorgonians were only observed in very rare instances and did not enter the dataset. Algae and Diadema antillarum Macroalgae constituted the most abundant algal functional group in the algal quadrats in one patch reef (TCI 5) and were codominant with crustose coralline algae in the other two (Table 3). In the SERs, turf algae predominated in nine, crustose coralline algae were predominant in eight, these two algal groups were approximately equally abundant in 471 six reefs and two had essentially equal abundances of all three algal groups. Thus, turfs and crustose corallines were about equally common in all but the shallow patch reefs, where macroalgae were comparatively abundant. Macroalgal heights averaged less than 1 cm in 1 8 reefs, and from 1 -2 cm in seven reefs (Table 3). By far the tallest algae (about 7 cm high) were found in a high-relief SER at Mouchoir Bank (TCI 1) where clumps of Turbinaha were seen to be overgrowing colonies of Montastraea spp. Macroalgal indices (absolute abundance of macroalgae x macroalgal height) were highest here and in the patch reefs (particularly TCI 5). There were no significant differences in crustose coralline algal abundance between patch reefs and either type of SER (ANOVA, F=1.14, p=0.335), but macroalgal abundance, turf algal abundance, macroalgal height and macroalgal index differed significantly between the patch reefs and the SERs, which did not differ from each other (ANOVAs for macroalgae, F=16.6, p<0.001; for turf algae, F=5.6, p=0.009; for macroalgal height, F=3.633, p=0.041; for macroalgal index, F=8.07, p=0.002). The abundance of macroalgae, turf algae and crustose coralline algae did not differ significantly between sheltered and moderately exposed SERs (t-tests, F=0, p=0.998 for macroalgae; F=0.52, p=0.477 for algal turfs F=0.29, p=0.594 for crustose coralline algae). However, macroalgal height and macroalgal index (a proxy for biomass) were significantly greater in moderately exposed SERs (height; 1.7 ± 2.2cm, index: 40.6 ± 79.5) than in the sheltered SERs (height: 0.6 ± 0.4 cm; index: 10.2 ± 15.9) (t-test, F=7.25, p"=0.013 for macroalgal height; F=7.54, p=0.01 1 for macroalgal index). No relationship was noted between either macroalgal height or macroalgal index and the number of stony coral recruits in the quadrats. No Diadema antillarum were found in any of the belt transects, nor elsewhere in the TCI reefs. DISCUSSION Notwithstanding the moderately low total cover by live stony corals, the reef ecosystems in the Turks and Caicos Islands gave the overall impression of being in good condition. The large amounts of standing dead stony corals in two of the patch reefs were clear evidence that the presently low cover of live stony corals reflected at least one previous mortality event. This was not the case on the SERs where the stony corals were in good health with a low prevalence of standing dead colonies (range 0-4%, n=25 reefs), hence their relatively low cover (usually <25%) may be a natural phenomenon. Since all the investigated reefs are within the influence of bank waters, it is possible that the latter exert a strong control over their scleractinian communities. Warmed or cooled bank waters spilling over the reefs may sufficiently stress scleractinians so as to preclude faster growth or higher recruitment. As the Turks and Caicos Islands are also situated within one of the 472 main hurricane paths, some control may also be exerted by high wave-energy events (e.g., Blanchon and Jones, 1997). The higher cover of live stony corals in the more exposed locations suggests greater influence by bank waters than by waves, which would be expected to produce the reverse pattern; however, it is alternatively possible that stony corals grow faster in windward reefs. Few healthy patches of A. palmata were present either in the surveyed patch reefs or in patches that were visited but not surveyed. (It may be that more could be found on the northern reefs between North Caicos and East Caicos; however, we were not able to survey this area.) For example, patch reefs built by dead or partly dead colonies of A. palmata cover extensive areas on the windward (eastern) side of Turks Bank and near Ambergris Cay in southeastern Caicos Bank. These Tong-dead’ colonies of^l. palmata may have been caused by diseases since intact skeletons were common. No information on the timing of death is available; however, it appears that xmny Acropora were already dead when surveyed by Sullivan et al. (1994). Some of the partially living colonies of^. palmata exhibited signs of what appeared to be infection by white-band disease. A measure of uncertainty as to the cause of the present die-back remains, however, since local fisheries and nature conservation authorities mentioned occurrences of fishing with chemicals (dish-washing liquid, possibly bleach or gasoline) in these patch reefs. Therefore, what we interpreted as white-band disease might rather have been recent mortality triggered by exposure to toxins. However, we saw no direct evidence of fishing with toxic substances. A. cervicornis was not seen in Mouchoir Bank which is likely to be an artifact of incomplete sampling. On the Caicos and Turks Banks, the small colonies of A. cervicornis possibly represented a new generation of recruits or survivors from a previous mortality event. By selective removal of A. cervicornis, previous outbreaks of disease could have contributed to the overall low cover of live stony corals. In contrast to the Cayman Islands, where large reef areas are covered by skeletons of A. cervicornis, no such skeletal remains were observed in the TCI. Had A. cervicornis been more common previously and killed by disease, its skeletons must have completely disappeared due to in-situ erosion or down-slope transport into deepwater, but given the persistence of its skeletons in the fossil record this scenario seems unlikely. The presence of white-band disease in some colonies of A. cervicornis is, however, evidence that acroporid diseases were present in the TCI. The absence of bleached stony corals and of 1 00% recently dead colonies are indications that no catastrophic mortality events had occurred shortly before our surveys were made. We thus presume that the mass bleaching event of 1998 had only minimal impact in the TCI. Similarly, the low count of standing dead colonies in the SERs indicates a similar absence of major mortality outbreaks in these deeper reef habitats for at least several previous years. The generally low rates of recent mortality (mean=3%) suggest that much of the reef system was in good condition overall. Nevertheless, in a quarter of the examined reefs the >10 cm stony corals exhibited moderately high rates of disease 473 (6.5-17%). Our surveys may have coincided with an outbreak of white plague that had disproportionately affected colonies of Montastraea faveolata. The paucity of Acropora palmata recruits (none encountered in the 1,41 1 quadrats) and the skewed size distribution of the >10 cm sized corals are suggestive of a pulse-like population replenishment by rare, high-recruitment events. The general skewness of the Montastraea distributions towards intermediate sizes may indicate that most of the colonies were of similar ages (resulting from a strong recruitment pulse), and/or be an indication of strong asexual recruitment by fragmentation or, less likely, that they simply do not grow very large in the TCI. Recruitment by small planulating scleractinians like Porites astreoides and Agaricia agaricites was an order of magnitude higher than by the larger, spatially dominant brooders (e.g., Acropora, Montastraea) in accord with general experience elsewhere in the wider Caribbean (e.g.. Smith, 1992). That macroalgal height was greater in the windward reefs than in the leeward reefs, notwithstanding their comparatively low hydrodynamic resistance compared to crustose coralline algae and algal turfs, is surprising. The overall scarcity of macroalgae, which accounted for <20% of the benthic algae in 75% (21/28) of the surveyed reefs is encouraging. However, our qualitative impression on Mouchoir Bank was one of unusually strong macroalgal overgrowth over dead stony corals. Whether this is a sign of degradation or a transient temporal phenomenon could only be verified with time-series data. High cover by macroalgae is generally seen as a sign of deteriorating reef quality (e.g., Steneck, 1994), in part because they restrict the recruitment of stony corals (Rogers et al., 1984). Coral planulae are thought to settle preferentially on crustose coralline algae (Johnson et al., 1991), hence rates of coral recruitment may be lower when crustose corallines are scarce. Although a high abundance of macroalgae did not always correlate with a low recruit count (Table 3), it was associated with some of the lowest recruitment observed on the Mouchoir Bank and near Ambergris Cay. However, it has to be noted that AGRRA sample sizes are not large enough to warrant detailed comparisons among sites or make a credible estimate of recruitment at any given site. In general we found that the reefs on Mouchoir Bank, which lacks any human population, were in worse condition than in our survey sites on the Turks and Caicos Banks, which was an unexpected result. The part of the Mouchoir Bank within Turks and Caicos Islands territory is difficult to police and protect and is the target of an intense, mostly illegal, fishery. Fishing vessels (reputedly mostly from the Dominican Republic) were reported to sometimes use fishing methods that are destructive to corals. Also, overharvesting of herbivores facilitates the expansion of macroalgae over corals. Remoteness from human population need not necessarily translate into “pristine” and “healthy” reefs. Rather, from our survey it appears that controlled use of reef resources near a moderately dense population may be more sustainable than largely uncontrolled activities in remote locations. 474 ACKNOWLEDGMENTS Funding for this survey was facilitated by the AGRRA Organizing Committee as described in the Forward to this volume, the Austrian Science Foundation (grant P-1 3 165-GEO) and the Marine Environmental Institute. Field support by the School for Field Studies, Arawak Inn, Big Blue Unlimited, Sandals Resorts and Fish Eye Diving is gratefully acknowledged. Logistical assistance from the Turks and Caicos Department of Fisheries and the Coastal Resources Administration and Management Project, is particularly appreciated. We thank AGRRA team member S.C.C. Steiner for local guidance. Special thanks to J. Lang whose insightful review greatly increased the manuscript's quality. Our sincere appreciation goes to R. N. Ginsburg for initiating the AGRRA process and to P.A. Kramer and P.R. Kramer for their tireless efforts in its development. R. Moyer is thanked for help with data processing. REFERENCES Antonius, A. 1995. Pathologic syndromes on reef corals; a review. Pp. 161-169. In: J. Geister and B. Lathuillere (eds.). Coral Reefs in the past, the Present and the Future. Proceedings of the Second European Regional Meeting, ISRS, Luxemburg, Publications du Service Geologique de Luxembourg, 29. Blanchon, P., and B. Jones 1997. Hurricane control on shelf-edge reef architecture around Grand Cayman. Sedimentology 44:479-506. Chiappone, M., K.M. Sullivan, and C. Lott 1996. Hermatypic scleractinian corals of the southeastern Bahamas: a comparison to western Atlantic reef systems. Caribbean Journal of Science 32:1-13. Ginsburg, R.N. (compiler) 1 994. Global Aspects of Coral Reefs: Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. 420 pp. Johnson, C.R., D. Muir, and A. Reysenbach 1991 . Characteristic bacteria associated with surfaces of coralline algae: a hypothesis for bacterial induction of marine invertebrate larvae. Marine Ecology Progress Series 74:281-294. 475 Lang, J., P. Alcolado, J.P. Carricart-Ganivet, M. Chiappone, A. Curran, P. Dustan, G. Gaudian, F. Geraldes, S. Gittings, R. Smith, W. Tunnell, and J. Wiener 1998. Status of coral reefs in the northern areas of the wider Caribbean. Pp. 123-134. In: C. Wilkinson (ed.), Status of Coral Reefs of the World - 1998. Australian Institute of Marine Science. Peters, E.C. 1997. Diseases of coral reef organisms. Pp. 1 14-139. In; C. Birkeland (ed.), Life and Death of Coral Reefs. C. Chapman and Hall, New York. Rogers, C.S., H.C. Fitz, M. Gilnack, J. Beets, and J. Hardin 1984. Scleractinian coral recruitment patterns at Salt River Canyon, St. Croix, U. S. Virgin Islands. Coral Reefs 3:69-76. Santavy, D.L., and E.C. Peters 1997. Microbial pests; coral disease in the western Atlantic. Proceedings of the Eighth International Coral Reef Symposium 1:607-612. Scholle, P.A., and N.P. James 1995. Marine Carbonates I: Models, Seismic Response and Quaternary of Florida/Bahamas. Society for Sedimentary Geology (SEPM), Photo CD 1 . Steiner, S.C.C. 1 999. Species presence and distribution of Scleractinia (Cnidaria: Anthozoa) from South Caicos, Turks and Caicos Islands. Bulletin of Marine Science 65: 861-871. Steneck, R.S. 1 994. Is herbivore loss more damaging to reefs than hurricanes? Case studies from two Caribbean reef systems (1978-1988). Pp. 220-226. In: R.N. Ginsburg (compiler) Global Aspects of Coral Reefs - Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Smith , S.R. 1 992. Patterns in coral recruitment and post-settlement mortality on Bermuda’s reefs: comparison to Caribbean and Pacific reefs. American Zoologist 32: 663-673. Sullivan K.M., M. Chiappone, and C. Lott 1994. Abundance patterns of stony corals on platform margin reefs of the Caicos Bank. Bahamas Journal of Science 1:2-12. Wanless H.R., and J.J. Dravis 1989. Carbonate Environments and Sequences of the Caicos Platform: Field Trip Guidebook T374. American Geophysical Union, Washington DC, 75 pp. Wells, S. (editor) 1988. Coral Reefs of the World, Volume 1: Atlantic and Eastern Pacific. United Nations Environment Program, Regional Seas Directories and Bibliographies. lUCN, Cambridge and UNEP, Nairobi, 373 pp. 476 Woodley, J.D., P. Alcolado, T. Austin, J. Barnes, R. Claro-Madruga, G. Ebanks-Petrie, R. Estrada, F. Geraldes, A. Glasspool, F. Homer, B. Luckhurst, E. Phillips, D. Shim, R. Smith, K. Sullivan-Sealy, M. Vega, J. Ward, and J. Wiener 2000. Status of coral reefs in the northern Caribbean and western Atlantic. Pp. 261-285. In: C. Wilkonson (ed.). Status of Coral Reefs of the World: 2000. Cape Ferguson, Queensland and Dampier, Western Australia. Australian Institute of Marine Science. Table 1. Site information for AGRRA coral and algal surveys in the Turks and Caicos Islands. 477 Site Name Site Latitude Longitude Survey Depth Benthic >10 cm stony % live stony coral code (°’N) (“'W) date (m) transects (#) corals (#/10 m) cover (mean ± sd) A. pal/nala patch reefs b/w Round and Gibbs Cay TC9 21 26.307 71 06.628 Aug 17 99 3.5 13 8 10.5 ±3.0 E.ofS. end of Grand Turk TCIO 21 27.534 71 06.818 Aug 17 99 2.5 13 7 8.0 ±4.0 Ambergns Cay 1 TC15 - - Aug 20 99 6.5 12 9 15.5±6.5 High relief shelf-edge reefs Lighthouse Point (anchor) TCI 21 31.139 71 08.037 Aug 14 99 17.5 10 11 16.0 ±9.0 North Point (anchor) TC2 21 32.220 71 06.553 Aug 14 99 16.0 13 10.5 16.5 ±6.0 Coral Garden TC3 21 27.493 71 09.301 Aug 14 99 11.5 11 14 25.5 ± 7.0 N. of Salt Cay (anchor) TC5 21 22.408 71 12.078 Aug 1 5 99 12.0 10 8.5 7.5 ±2.0 Mouchoir Bank TCIl 20 59.159 70 47.008 Aug 18 99 22.5 11 11.5 10.5 ±4.0 The Arch TC13 21 28.996 71 31.062 Aug 19 99 10.5 10 12.5 20.5 ±5.0 Ambergris Cay 2 TC16 21 22.359 71 35.949 Aug 20 99 20.0 12 10 12.5 ±8.5 The Pinnacles (Grace Bay) TC23 21 48.841 72 11.219 Aug 26 99 10.0 10 12.5 31.0±9.5 Football Field TC26 21 54.381 72 06.916 Aug 27 99 19.0 8 16.5 28.5 ±7.5 Low-relief shelf-edge reefs West of Little Sand Key TC4 21 23.658 71 10.088 Aug 1 5 99 9.5 10 10.5 14.5 ±6.0 Casey's Wall (anchor, W. of TC6 21 18.201 71 13.388 Aug 16 99 11.5 9 14.5 24.5 ±6.5 Salt Cay) Black Forest TC7 21 28.754 71 09.211 Aug 16 99 11.5 9 13 21.0±4.5 Chief Minister's House TC8 21 26.443 71 09.294 Aug 16 99 16.5 9 12.5 24.0 ±8.5 Mouchoir Bank TC12 21 01.369 70 49.108 Aug 18 99 13.5 11 14 17.0±6.0 Airplane TC14 21 32.762 71 27.348 Aug 19 99 16.5 10 11.5 13.5±3.5 (No Name) TC17 21 28.029 71 33.125 Aug 21 99 15.5 11 11 14.5 ±2.0 Fish Hole TC18 21 29.072 7130.629 Aug 21 99 13.0 11 10.5 12.0 ±3.0 French Cay TC19 21 29.357 72 13.456 Aug 23 99 14.5 10 0 22.0 ±6.0 West Sand Spit TC20 21 23.285 72 08.637 Aug 23 99 13.5 10 12 20.0 ± 10.0 Spanish Anchor (West TC21 21 38.739 72 28.473 Aug 24 99 15.5 11 8 12.5 ±4.0 Caicos wall) West Caicos Wall-middle TC22 21 39.890 72 28.211 Aug 24 99 18.5 9 12 22.0 ±4.5 Coral Gables (North side of TC24 21 49.080 72 11.010 Aug 26 99 11.0 8 12.5 22.0 ±8.5 Provo) Grace Bay (North of TC24) TC25 21 49.809 72 10.395 Aug 26 99 10.0 10 10.5 14.0 ±6.0 Grouper Hole TC27 - -- Aug 27 99 11.5 8 15 31.0±7.0 Aquarium West TC28 21 48.518 72 13.542 Aug 27 99 17.5 10 11 12.5 ±3.5 All sites (mean ± sd) 1.3 11. 1 18.0 ± 5.5 478 Table 2. Size and condition (mean ± standard deviation) of all stony corals (>10 cm diameter) by site in the Turks and Caicos Islands. Site naiiie/Site code Stony corals Partial-colony mortality (%) Stony corals (%) n Diameter Height Recent Old Total Standing Bleached Diseased (cm) (cm) dead A. palmala patch reefs b/w Round and Gibbs Cay/TC9 107 103.0 ±95.0 66.0 ±74.0 1.0±7.5 47.0 ±40.5 61.5 ±36.0 20.5 0 4.5 E.ofS. end of Grand Turk/TClO 91 78.5 ±77.5 54.0 ±57.0 <0.5 ±0.5 41.0 ±40.0 56.5 ±36.0 15,5 0 4.5 Ambergris Cay 1/TC15 107 55. 5± 50.5 33.0 ±27.5 3.0± 12.5 21.0±28.0 34.0 ±29.5 3 0 7 High relief shelf-edge reefs Lighthouse Point (anchor )/TCl 112 29.0 ± 17.5 16.5 ± 14.0 0.5 ±2.0 18.5 ±23.5 29.5 ±2 3.5 0 0 2 North Point (anchor)/TC2 136 37.5 ±26.5 20.0 ± 15.0 1.5 ±7.0 12.5 ± 16.5 22.0 ± 18.0 0 0 2 Coral Garden/TC3 154 40.5 ±24.0 26.0 ± 16.5 5.5± 18.5 21 5 ±2 4.0 34.0 ±27.5 2.5 0 3 N. of Salt Cay (anchor)5/TC5 84 26.0 ± 14.5 18 5 ± 16.5 1.0 ±6.0 20.0 ± 27.0 31.5 ±28.0 1 0 3 Mouchoir Bank/TCl 1 126 31.5±1 4.0 26.5 ± 14.5 4.0± 9.5 31.5 ±26.5 41.5±26.0 4 0 7 The ArclVTCI3 123 44.0 ±28,0 29.0 ±20.5 L5±4.5 25.0 ±25.0 32.0 ±24. 1 0 4 Ambergris Cay 2/TC16 117 32.0 ±20.0 21. 5± 15.0 4.5 ± 13.5 15,0 ±20.0 27.5 ±23.5 0 0 1 The Pinnacles (Grace Bay)/TC23) 127 50.0 ±41.5 39.0 ±38.5 2.5 ±9.5 16.5 ± 19.0 24.5±21.5 0 0 2 Football Field/TC26 132 48.0 ±29.0 28.0 ± 17.0 3.5± 1.0 18.5 ±23.5 30.0 ±25.5 1.5 0 13 Low-relief shelf-edge reefs West of Little Sand Cay/TC4 105 37.5 ±22.0 22.0 ± 17.5 3.0 ± 10.5 27.5 ±29.0 38.5 ±31.5 2 0 1 Casey's Wall (anchor, W. of Salt 132 40.5 ±2 6.5 25.5 ± 19.0 1.0±4.5 21.5 ±26.0 32.0 ±25.5 0 0 2 Cay)/TC6 Black Forest/TC7 118 40.5 ±1 9.0 27.5 ± 18.0 2.0± 11.0 27.0 ±24.5 33.5 ±25.0 0 0 2 Chief Minister's House/TC8 111 57.0 ±44.0 37.0 ±28.0 4.9 ± 13.0 22.5 ±24,5 34.0 ±26.5 2 0 1 Mouchoir Bank/TCl 2 153 53.0±41.5 33.5 ±23.5 6.0 ± 13.0 29.0 ±29.5 43.5 ±30.5 2.5 0 17 Airplane/TC14 116 38.0 ±24.0 27.0 ± 18.5 2.5 ±9.5 34.0±31.0 45.0 ±30.5 3.5 0 4 (NoName)/TC17 123 37.5 ± 19.0 25.0 ± 14.0 2.5 ± 10.0 31.5 ±3 1.0 41.0 ±30.0 1.5 0 4 Ftsh Hole/TC18 117 33.0 ±18.5 21. 5± 14.0 7.5 ± 17.5 21.5 ±24.5 36.5 ±29.0 1.5 0 2 French Cay/TC19 128 43,5 ±23.5 30.5 ±20.5 2.5 ± 10.5 18.5 ±23.0 31.0±25.5 0 0 2 West Sand Spit/TC20 118 42.0 ±29.5 31 0±21.5 1.0 ±4,0 17.0 ±22.0 25.5 ±22.5 0 0 3 Spanish Anchor (West Caicos wall) 90 45.0 ±33.0 32.5 ±24.5 3.5 ± 12.5 22.0 ±27.0 34.5 ±29.0 2 0 9 /TC2I West Caicos Wall-middle/TC22 110 59.0 ±47.5 36.5 ±33.5 3.0 ±9.5 20.5 ±26.0 36.0 ±29.5 3.5 0 9 Coral Gables (North side of Provo) 101 49.0 ±33.5 33.0 ±28.5 3.5 ± 12.5 16.5 ±22.0 26.0 ±25.0 1 0 4 /TC24 Grace Bay (North of TC24)/TC25 106 39.5 ±21.0 30.5 ± 19.0 L5± 10.0 22.5 ±26.5 35.0 ±27.0 0 0 6 Grouper Hole/TC27 121 51.5 ±39.0 37.5 ±28.5 2.5 ± 10.0 19.5 ±23.5 29.0 ±26.5 1 0 13 Aquarium West/TC28 111 32.5 ± 15.0 22.0 ± 14,0 2.5± 10.0 17.0 ±23.0 29.0 ±26.0 2.5 0 6 All sites (mean ± sd) 119 45.2±38.7 30.1 ±28.2 2.8 ± 10.7 23.2 ±27.1 34.8 ± 28.5 2.5 0 5 479 Table 3. Algal characteristics, density of stony coral recruits and Diadema antillarum (mean ± standard deviation) by site in the Turks and Caicos Islands. Site name Quadrats Absolute abundance (%) Macroalgal Recruits Diadema (#) Macroalgae Turf Crustose Height Index' (#/.0625 m^) (#100 m^) algae coralline algae (cm) A. palmata patch reefs b/w Round and Gibbs Cay/TC9 55 45.0 ±35.0 28.5 ±32.0 51.0±36,5 2.2 ± 1.2 92 0.02±0.1 0 E.ofS. end of Grand Turk/TClO 60 38.0 ±30.5 26.5±23.0 36.0 ±24.5 1.8± 1.2 71 0.2±0.5 0 Ambergris Cay 1/TCI5 60 62.0 ±25.5 14.0 ±20.5 24.0 ± 16.0 3.8 ±2.8 235 0.2±0.6 0 High relief shelf-edge reefs Lighthouse Point (anchor)/TCl 45 16.5 ± 19.0 28.0 ±25.5 55.5 ±26.5 1.0± 1.0 16 0.3±0.6 0 North Point (anchor)/TC2 65 26.0 ± 17.5 22.5 ±25.0 51.5±23.5 2.0 ±3.2 49 0.1±0.3 0 Coral Garden/TC3 55 4.0± 11.5 51.0 ±28.5 45.5 ±26.5 0.2 ± 0.6 <1 0.4±0.6 0 N. of Salt Cay (anchor)/TC5 50 17.0 ± 19.5 34.0 ±23.5 48.5 ±27.0 0.8 ±0.8 13 0.3±0.6 0 Mouchoir Bank/TCI 1 55 33.5 ±25.5 32.0 ±23.0 34.0 ± 20.0 7.0 ± 5.4 233 0.2±0.5 0 The Arch/TCI 3 50 2.5 ±7.0 52.5 ±25.5 45.0 ±25.0 0.2 ±0.6 <1 0.6±1.3 0 Ambergris Cay 2/TC 1 6 60 11. 5± 16.5 24.0 ±22.0 64.5 ±27.5 1.6 ±3.0 17 0,3±0.7 0 The Pinnacles (Grace Bay/TC23) 50 0.5 ±3.0 58.0±21.0 41.5±21.0 0.2 ±0.2 <1 0.2±0.5 0 Football Field/TC26 40 39.0 ±35.0 28.5 ±29.5 32.5 ±27.0 1.2± 1.0 43 0.2±0.5 0 Low-relief shelf-edge reefs West of Little Sand Key/TC4 50 8.5 ± 13.5 49.5 ±24.5 42.5 ±27.0 0.6 ±0.8 5 0.5±0.8 0 Casey's Wall (anchor, W. of Salt 45 4.0 ±8.0 53,5 ±22.0 43.0 ± 19.0 0.6 ± 1.0 2 0.2±0.5 0 Cay)/TC6 Black Forest/TC7 45 2.0 ±6.5 79.5 ± 18.5 18.5 ± 18.5 0.2 ±0.8 ,5 0.4±0.7 0 Chief Minister's House/TCS 41 1,0 ±2.5 84.5 ± 19.5 34.5 ±32.5 0.8 ±2.4 <1 0.3±0,5 0 Mouchoir Bank/TC 1 2 55 11.0± 19.5 54.0 ± 19.0 35.0 ± 19.0 1.8±3.6 20 0.1±0.4 0 Airplane/TC14 47 6.5 ±13.5 36.5 ±22.0 57.0 ±22.0 0.6 ± 1.0 4 0.5±0.9 0 (No Name) /TCI 7 55 L5±6.5 46.5 ± 19.5 52.0 ±20.5 0.2 ±0.6 <1 0.3±0.6 0 Fish Hole/TC18 54 2.5 ±6.5 49.0 ±27.5 48.5 ±26.5 0.4 ±0.8 <1 0.4±0.8 0 French Cay/TCI9 50 37.5 ±25.0 15.5±21.0 47.5 ±28.5 1.6± 1.0 56 0.2±0.5 0 West Sand Spit/TC20 50 13.5 ±22.0 47.5 ±30.5 39.0 ±26.0 0.8 ± 1.0 9 0.3±0.7 0 Spanish Anchor (West Caicos wall) 49 13.5 ±22.5 35.5 ±20.0 51.0 ± 19.5 0.8 ± 1.0 11 0.4±0.6 0 /TC21 West Caicos Wall-middle/TC22 45 14.5 ± 17.5 45.0 ±24.0 40.0 ± 19.5 0.8 ±0.8 12 0.4±0.7 0 Coral Gables (North side of Provo) 40 l.0±4.0 55.5 ±23.5 43.0 ±22.5 0.2 ±0.4 <1 0.6±0.9 0 /TC24 Grace Bay (North of TC24)/TC25 50 <0,5 ±0.5 62.0 ±24.0 38.0 ±24.0 <0.1 ±0.2 <1 0.5±0.6 0 Grouper Hole/TC27 40 1.5±4.0 66.0 ±23.5 33.0 ±23.5 0.2 ±0.6 <1 0.8±1 0 Aquanum West/TC28 50 8.5±11.5 39.5 ±29.5 52.5 ±25.0 0.6 ±0.8 5 0.4±0.7 0 All sites (mean ± sd) 50.4 18.4±28.3 52.2 ±35.1 55.7 ± 35.9 1.2± 2.4 21.82 0.3±0.7 'Macroaalgal index = absolute macroalgal abundance x macroalgal height 480 Figure 1. AGRRA survey sites in the Turks and Caicos Islands. See Table 1 for site codes. Wind rose for the southern Bahamas, from R.N. Ginsburg in RA. Scholle, and N.R James (1995). ASSESSMENT OF THE CORAL REEFS OF THE TURKS AND CAICOS ISLANDS (PART 2: FISH COMMUNITIES) BY KAHO HOSHINO,’ MARILYN BRANDT,^ CARRIE MANFRINO/ BERNHARD RIEGL,^ and SASCHA C. C. STEINER^ ABSTRACT Ecologically and commercially significant coral reef fishes were surveyed at 28 sites in the Turks and Caicos Islands during August 1999. Our results constitute the first quantitative census of these fishes and can serve as baseline information for subsequent studies. Their density and size generally were highest off West Caicos and lowest in Mouchoir Bank. Herbivore density overall showed no correlation with macroalgal index (a proxy for biomass) or live stony coral cover, but surgeonfish density was positively correlated with macroalgal index. Species richness of these select fishes was positively correlated with the species richness of stony corals that were >10 cm in diameter. Current fishing pressures overall were low, and the reef-fish communities appeared relatively intact on the Turks and Caicos Banks. However, overfishing and destructive fishing practices have negatively impacted the reef fish communities on Mouchoir Bank. ' Formerly Bren School of Environmental Science and Management, University of California, Santa Barbara, Santa Barbara, CA 93106-5131. National Center for Caribbean Coral Reef Research, Rosenstiel School of Marine and Atmospheric Science, 4600 Rickenbacker Causeway, Miami, FL 33149-1098. Kean University, Department of Geology and Meteorology, 1000 Morris Ave. Union, NJ, 07083. National Coral Reef Institute, Oceanographic Center, Nova Southeastern University, 8000 N. Ocean Drive, Dania, FL 33004 and Institute for Geology and Paleontology, Karl-Franzens-University Graz, 8020 Graz, Austria. Email: rieglb@nova.edu ^ Institute for Tropical Marine Ecology Inc. P.O. Box 944, Roseau, Commonwealth of Dominica, West Indies. 482 INTRODUCTION Fish are known to occur at high species diversity and density in undisturbed coral reefs. Although their exact numbers are unknown (Sorokin, 1993), they play important roles in reef ecosystems as herbivores and as top predators (e.g., Hatcher and Larkum, 1983; Hay, 1997). Reef fishes can be sensitive indicators of general reef condition (Pattengill-Semmens and Semmens, this volume). Furthermore, certain large-sized predators such as groupers (serranids) have been historically targeted by commercial as well as sport fishers and they are known to be particularly susceptible to fishing pressure (Russ, 1991; Sluka et ah, 1998). Therefore, their relative abundance should serve as a good indicator of fishing pressures. The Turks and Caicos Islands (TCI), which comprise the southeastern extent of the Bahamian Archipelago, lie between 21° and 22° N latitude and 71° and 72° 30' W longitude (Fig. 1). There are seven inhabited islands, one uninhabited island, and approximately 40 low-lying coral limestone cays distributed on two banks (Turks and Caicos). Part of the entirely submerged Mouchoir Bank, which is located approximately 65 km southeast of Turk Bank between the Mouchoir Passage and the Dominican Republic, also belongs to the TCI. The prevailing easterly trade winds keep sea conditions on the eastern sides of the banks and islands choppy, while the western sides are generally calm. Currents generally run to the north. Over 300 km of coral reef are reported to exist in the TCI (Wells, 1988; Turks and Caicos National Parks Advisory Committee, 1998, report). The TCTs coral reefs are essential resources for tourism and provide seafood for local consumption. Fisheries, often directly reef-related, are economically important with lobster and conch being the only national exports (Woodley et ah, 2000). The tourism industry has developed rapidly during the past decade, particularly for recreational diving. Increased development, population increase, and illegal fishing may in future be challenges for the integrity of the TCTs coral reefs; however, at present they are still remote from intense human impacts. In 1975, a National Parks Ordinance established 19 marine protected areas with three levels of protection (Woodley et al., 2000) which collectively cover about 700 km . Active management is currently restricted to protected areas near Providenciales, where much of the TCTs populace resides (Woodley et al., 2000). Nevertheless, given its total land area of 505 km^, the TCI government has legislated proportionately more protected areas than any other country in the world. In addition, the government has established a self-financing system called the Conservation Fund to ensure long-term sustainability of the national park system. The major objective of this study was to document fish community composition and structure in the TCI in a snapshot fashion with the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocols. Because of their remote locality, these islands afforded 483 an opportunity to study relatively “pristine” reef fish communities with few anthropogenic impacts. Relative to many other tropical western Atlantic ecosystems, the reef communities in the TCI are understudied (Chiappone et ah, 1996). Our results represent the first systematic census of key fish species and provide baseline data for future monitoring and research in the TCI. METHODS A total of 28 sites (Fig. 1) were studied: two on Mouchoir Bank; 10 on Turks Bank; and 16 in three major localities on the larger Caicos Bank (4 near West Caicos and French Cay, 6 near Providenciales, and 6 near South Caicos). We selected sites that were representative of special interests (i.e., reported to be heavily impacted or of touristic, fisheries and/or conservation value) and strategically accessible (at established dive sites with mooring buoys). An effort was made to space the sites as evenly as possible within all available exposures and reef types, but the small size of the dive boats and prevailing sea conditions restricted surveys to areas of moderate exposure and/or short traveling distance. Although the northern side of Caicos Bank and much of Mouchoir Bank were not exhaustively investigated, a relatively even mix of exposed and sheltered reefs was obtained. Three shallow (<8 m) patch reefs dominated by Acropora palmata were sampled (TC9, TCIO, TCI); however, a continuous Acropora palmata reef crest was not observed in the surveyed areas. The remaining 25 reefs were located on the outer margin of the upper shelf-edge reefs in low-relief (<5 m, n=16) and high-relief (>5 m. n=9) habitats with relatively dense coral growth (Table 1). The surveys were carried out following the AGRRA Version 2.2 fish protocol (see Appendix One, this volume) by a team of four divers who generally spent between one and two hours underwater at each site. Each belt transect was 30 m x 2 m. Counts of groupers were restricted to species of Epinephelus and Mycteroperca; scarids (parrotfishes) and haemulids (grunts) less than 5 cm in length were not tallied due to the difficulty in accurate identification. Each diver also performed one 20-minute-long Roving Diver Technique (RTD) survey at every site. One day of consistency training using plastic fish models of varying sizes (10 cm, 20 cm, 40 cm) was completed prior to initiating the surveys. Field identifications were based on Humann (1994). While we attempted to identify all species as accurately as possible, some errors are likely to have occurred, particularly with the smaller parrotfishes and grunts. Fish in these size classes were not very common, however, and are believed to have introduced little bias into the dataset. The belt-transect data were used to compare the mean density (as numbers/ 100 m ) and size distribution (in cm) of selected fishes among the five geographic areas using Analysis of Variance (ANOVA). The relationship between key herbivore density and 484 macroalgal index (absolute macroalgal abundance x macroalgal height) was analyzed using Kendall's rank correlation test, as was the relationship between total AGRRA fish diversity and the diversity of stony corals >10 cm in diameter (select stony corals). Relationships between fish density and other benthic habitat variables (live stony coral cover; height, diameter and mortality of select stony corals; depth) were investigated using linear and multiple regressions. (The AGRRA benthic survey data are summarized in Riegl et al., this volume.) Whenever necessary, the data were transformed prior to performing a parametric statistical analysis (e.g., log-transformation was used for fish size, square-root transformation for fish density, and arcsine transformation for percentile data). RESULTS Belt Transects Species richness. A total of 4,597 fishes belonging to 46 of the species in the AGRRA fish list, including seven each of groupers and parrotfishes, were counted in 279 belt transects (Tables 1, 2). Mean AGRRA species richness/site in the five geographic locations varied from 17 in Mouchoir to 25 in West Caicos, averaging 22 for the TCI as a whole (Fig. 2). The West Caicos Wall (site TC22) had the highest record (30 species) for species richness (Table 1). Pairwise comparisons (Mann- Whitney U-test) revealed that the species richness of the AGRRA fishes in Mouchoir Bank was significantly lower than in Turks Bank (p=0.0417), Providenciales (p=0.0318), and West Caicos (p=0.0301). Moreover, the AGRRA fish species richness in the belt transects was positively correlated (Fig. 3; Kendall's rank correlation, 0.018m; n=25) reefs in Figure 4. Angelfish (pomacanthids), butterflyfish (chaetodontids) and triggerfish (balistids) were significantly more abundant in the deeper reefs (ANOVA, p<0.05), whereas surgeonfish (acanthurid) densities were significantly higher in the shallow reefs (ANOVA, p<0.05). 485 Figure 2. Species richness (mean/site ± standard deviation) of AGRRA fishes for the Turks and Caicos Islands (TCI, 4,597 fishes, 28 sites) and five geographic areas: Turk Bank (1,524 fishes, 10 sites). South Caicos (972 fishes, 6 sites), Providenciales (1,198 fishes, 6 sites). West Caicos (691 fishes, 4 sites), Mouchoir Bank (275 fishes, 2 sites). Figure 3. Correlation plot between species richness of all stony corals (>10 cm diameter) and all AGRRA (belt-transect) fishes by site in the Turks and Caicos Islands. 486 □ Shallow (<8m) ■ Deep (>8]n) Fish Family Figure 4. Mean fish abundance (no. individuals/100 ± sd) for AGRRA fishes at <8m (629 fishes, 3 sites) and >8 m (n=3,968 fishes, 25 sites) in the Turks and Caicos Islands. * = statistically significant difference (ANOVA, p<0.05) between shallow and deep sites. The average density of the select fishes in each of these eight families in the TCI as a whole and in the five geographic locations (Table 3 A) are shown in Figure 5. A statistically significant difference among geographic areas was found in mean snapper (lutjanid) density (ANOVA, P=0.0234), with the highest value occurring in the reefs around West Caicos. Mean densities of grunts, snappers and triggerfishes were significantly lower (Mann- Whitney U-test, p=0.0405, 0.0246, 0.0405 respectively) in Mouchoir Bank than means for the pooled data in the remaining TCI reefs (Fig. 6A). The average density for five of these families in each reef is given in Table 4. There were no significant differences in the densities of parrotfish [Kruskal-Wallis (K- W) H Test p=0.16], snappers (K-W H-Test, p=0.91), and grunts (K-W H-Test, p=0.19) among the three types of investigated reefs, but significant differences were found in the densities of surgeonfishes (K-W H Test, p=0.02) and groupers (K-W H Test, p=0.028). Parrotfish and surgeonfish densities were higher in the shallow palmata patch reefs than in the deeper reefs. Groupers had higher densities in high-relief shelf-edge reefs. Size. The size (as total length) distributions of herbivores (parrotfish >5 cm, surgeonfish, the yellowtail damselfish Microspathodon chrysurus) and carnivores (snapper, select grouper) at shallow (<8 m) and deeper (>8 m) reefs are summarized in Figure 7. Shallow and deeper reefs differed significantly for both of these important trophic groups (G-test, both p<0.001) with deeper reefs generally having larger fishes. Average sizes were higher in carnivores than herbivores at both depths even though the modes for both groups were in the 11-20 cm-size class in the deeper reefs. 487 O O % 5 C/5 c 5 cm and surgeonfish) nor surveyed parrotfish density was correlated with macroalgal index (Fig. 8 A, B; Kendall’s rank correlation tests, p>0.1). However, surgeonfish density was positively correlated with macroalgal index (Fig. 8C; Kendall’s test, 0.010.1). Neither the average density of the key herbivores nor the average total fish density showed a significant relationship with percent live stony coral cover (linear regression, p>0.1. Fig. 9 A, B). Roving Diver Surveys The total number of species observed during the RTD surveys was 120. The 25 most commonly seen species, along with their sighting frequency in the REEF (2000) database, are shown in Table 5. The highest total species richness occurred in West Caicos and Providenciales, while the lowest was in Mouchoir Bank. Parrotfishes were the A Carnivores 70.0 60.0 ^ 50.0 ^ 40.0 e 3 30.0 O” ti. 20.0 u< 10.0 0.0 59.4 □ Shallow (<8m) ■ Deep (>8m) 51.6 21-30 31-40 Size "ategory (cm) B Herbivores w s a 3 O" t 30. Size category (cm) Figure 7. Size frequency distribution of (A) carnivores (lutjanids, select serranids) and (B) herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) at <8m (629 fishes, 3 sites) and >8 m (3,968 fishes, 25 sites) in the Turks and Caicos Islands. 490 most common of the AGRRA fishes seen in the Turks and Caicos. Butterflyfish and angelfish were the least commonly encountered, with the exception of Mouchoir Bank where grunts and snappers were seen least frequently. a. • N s a 30 25 20 If 10 5 0 w Kenddls Earik Correkti»i,p>0 .1 .Q .1 n 1 a ^ b. Log (Macroalgal Index) llO 14 o o 12 10 9 a dj 6 T3 4 c cd o 2 c/5 • ♦ i*-' - ' ♦♦ * 4. ^ •4 Kendall's Rank Correlation, p>0.1 •3-2-10 12 3 Log (Macroalgal Index) Log (Macroalgal Index) Figure 8. Correlation plots between mean log macroalgal index and mean density (no. individuals/100 m^) of (A) herbivores (acanthurids, scarids >5 cm), (B) scarids >5 cm, and (C) acanthurids by site in the Turks and Caicos Islands. 491 Figure 9. Relationship between mean live stony coral cover (%, arcsin-squareroot transformed) and mean density (no. individuals/100 m^, squareroot transformed) of (A) Key AGRRA herbivovres and (A) all AGRRA fishes, by site in the Turks and Caicos Islands. DISCUSSION The average grouper density in the TCI (1 .62/100 m , Table 3A) was higher than Sluka et al. (1998) had found in the Florida Keys (0.01-0.13/100 m^) and the Exuma Cays, Bahamas (0.10-0.90/100 m ). Although comparisons must be made with caution due to the different methodologies used, it would appear that current fishing pressures on the Turks and Caicos Banks are relatively low. That groupers had higher densities in high-relief shelf-edge reefs may reflect some correlation between density and habitat complexity. The large numbers of grunts, particularly in Aquarium West (TC 28), were mainly small individuals and likely to have resulted from active recruitment. The average length of groupers in the TCI was 17.5 cm. While smaller species like coney (Epinephelus fulvus) and graysby (E. cruentatus) were the most abundant, larger species such as Nassau (Epinephelus striatus) and black (Mycteroperca bonaci) 492 groupers were also recorded in the belt transects in eight reefs. Only two percent (6/271) of the groupers in the belt transects were larger than 40 cm in length. Equally large groupers were also sighted during several dives outside our belt transect areas. Since much of the area, whether protected or unprotected, is subject to some spearfishing, it is likely that large groupers are either so shy or so rare as to be seldom seen (Woodley et ah, 2000, refer to “uncontrolled fishing in the marine parks” of the TCI.) Illegal poaching appeared to be prevalent on Mouchoir Bank, probably due to its remote locality. We observed the use of spearguns. We deduced possible remnants of destructive fishing practices, such as chemicals and dynamite, from the peculiar way that stony corals had lost tissues and the way rubble from broken colonies was strewn on the reefs. Supporting information of such abuses was verbally provided by local fishery patrol officers. Notwithstanding our small sample size [two sites (TCI 1 and TCI 2), 20 transects], significantly lower species richness for the AGRRA fishes (Table 1), plus significantly lower densities for three of these families and smaller sizes for two families, of the AGRRA fishes (Fig. 6A,B), were recorded here. In summary, the Mouchoir Bank reefs appeared to suffer from overfishing and destructive fishing practices. The number of species found during the RDT surveys (120) was much lower than the total number (319 as of May 2000) reported from the TCI in the Reef Environmental Education Foundation (REEF) database in 2002. No new species records resulted from our surveys; however, with a more extended sampling effort, it is likely that some new records could have been obtained. The AGRRA belt transects are focused on ecologically and commercially important fishes to enhance the efficiency of the surveys; therefore, the data only provide species richness and diversity within a selected group of species. The RDT census provides supplementary data on the entire fish assemblage. The number of species recorded can differ substantially depending on the total number of divers, their level of experience, and the search time as seen in the REEF survey reports (at www.reef org) which clearly show different results between expert and novice divers. Since the objectives of the AGRRA project is to obtain a regional perspective on the status of Caribbean coral reefs based on a standardized methodology, the results of the RDT surveys should be treated with some caution. The overall results from the 1 999 AGRRA survey revealed that the TCI reef fish communities are relatively diverse and healthy, except on Mouchoir Bank. Although reefs accessible from Providenciales (i.e., Providenciales, West Caicos) receive the largest number of snorkelers and scuba divers, currently their fish communities appear to be in relatively good condition. This may be due to the fact that most of the area is designated as marine parks in which fishing is prohibited. Similar designations are either lacking or not well enforced in South Caicos and Turk, which historically have had higher levels of fishing activity. That fish density and diversity showed nearly significant positive relationships with the size (as diameter and height, proxies for substratum 493 architectural complexity) of the >10 cm stony corals is in general agreement with previous findings by Ault and Johnson (1998a, 1998b) and Chabanet et al. (1997). The relationships found in the TCI between the diversity of these stony corals and the diversity of ecologically and commercially important fishes are rarely addressed in the literature (but see Chabanet et al., 1997). Hence the AGRRA project will provide an excellent opportunity for continued exploration of this topic. ACKNOWLEDGMENTS The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. The Gilo Family Foundation and Reef Environmental Education Foundation provided funds for this research. B. Riegl was also supported by Austrian Science Foundation Grant PI 3 165-GEO. We acknowledge support by Sea Eye Divers, Big Blue Unlimited, Arawak Inn and Beach Club, South Caicos Ocean Haven, the School for Field Studies on South Caicos Island, Beaches Resort and Spa, Turks and Caicos Department of Tourism, Department of Environment and Coastal Resources, Department of Police, and Inter-Island Airways. We thank our seientifie party for their hard work in the field. REFERENCES Ault, J.R., and C.R. Johnson 1998a. Spatially and temporally predictable fish communities on coral reefs. Ecological Monograph 68:25-50. Ault, J.R., and C.R. Johnson 1998b. Spatial variation in fish species riehness on coral reefs: habitat fragmentation and stochastic structuring processes. Oikos 82:354-364. Chabanet, R, H. Ralambondrainy, M. Amanieu, G. Faure, and R. Galzin 1997. Relationships between coral reef substrata and fish. Coral Reefs 16:93-102. Chiappone, M., K.M. Sullivan, and C. Lott 1996. Hermatypic scleractinian corals of the southeastern Bahamas: A eomparison to western Atlantic reef system. Caribbean Journal of Science 321-13. Hatcher, B.G., and A.W.D. Larkum 1983. An experimental-analysis of factors controlling the standing crop of the epilithic algal community on a coral reef. Journal of Experimental Marine Biology and Ecology 69:61-84. 494 Fish-seaweed interactions on coral reefs: effects of herbivorous fishes and adaptations of their prey. Pp. 96-119. In: C.H. Birkeland, (ed.). Life and Death of Coral Reefs. Chapman & Hall, New York. Reef Environmental Education Foundation. World Wide Web electronic publication, www.reef.org/ Coral reef fisheries: effects and yields. Pp. 601-635. In: P.F. Sale (ed.), The Ecology of Fishes on Coral Reefs. Academic Press, New York. Sluka R., M. Chiappone, K.M. Sullivan, T.A. Potts, J.M. Levy, E.F. Schmitt, and G. Meester 1998. Density, species and size distribution of groupers (Serranidae) in three habitats at Elbow Reef, Florida Keys. Bulletin of Marine Science 62:219-228. Sorokin, Y.I. 1993. Coral Reef Ecology. Springer- Verlag, Berlin, 465 pp. Wells, S. (editor) 1988. Coral reefs of the world, Volume 1: Atlantic and eastern Pacific. United Nations Environment Program, Regional Seas Directories and Bibliographies. lUCN, Cambridge and UNEP, Nairobi, 373 pp. Woodley, J.D., P. Alcolado, T. Austin, J. Barnes, R. Claro-Madruga, G. Ebanks-Petrie, R. Estrada, F. Geraldes, A. Glasspool, F. Homer, B. Luckhurst, E. Phillips, D. Shim, R. Smith, K. Sullivan-Sealy, M. Vega, J. Ward, and J. Wiener 2000. Status of coral reefs in the northern Caribbean and western Atlantic. Pp: 261-285. In: C. Wilkonson (ed.). Status of Coral Reefs of the World: 2000. Cape Ferguson, Queensland and Dampier, Western Australia. Australian Institute of Marine Science. Hay, M.E. 1997. REEF 2000. Russ, G.R. 1991. 495 to TJ G cd CO O _o 'S u T3 G cd CO l-H od On -H -H tT) IT) Tt oo — H On m 00 (N (N (N z; ^ y y y y u u H H H H ^ f-. VC m NO — (N o Tf 2 o m in o — — ^ o -H 00 ^ o o VC ^ O O m in m in rn ^ cn in 3 < < < 3 < — rf — On ^cm) Haemulidae (>^cm) Lutjanidae Serranidae' A. palmata patch reefs TC9 13.33 16.00 7.50 7.33 0.50 TCIO 18.83 8.83 0.33 0.83 0.83 TC15 13.00 9.50 0.17 1.33 0.67 Mean ±sd 15±3.2 11.4±3.9 2.7±4.2 3.2±3.6 05±0.83 High-relief shelf-edge reefs TCI 6.00 4.50 4.00 0.17 1.00 TC2 7.64 6.25 1.53 0.56 0.97 TC3 2.33 11.50 13.00 8.00 0.83 TC5 6.00 4.50 4.00 0.17 1.00 TCll 6.67 7.83 0 0 1.50 TC13 4.67 11.67 5.17 2.50 1.83 TC16 3.83 13.17 6.50 2.33 1.33 TC23 6.33 12.00 3.50 7.00 2.17 TC26 4.17 9.50 4.33 3.83 0.67 Mean ±sd 5.2±1.6 8.9±3.3 4.7±3.6 2.7±3.01 0.7±2.2 Low relief shelf-edge reefs TC4 9.17 5.33 0.33 0 1.33 TC6 6.83 6.33 1.83 0 2.33 TC7 4.83 8.67 0.50 1.67 1.67 TC8 2.00 9.83 3.17 6.50 2.33 TC12 9.33 8.83 0.33 0.17 2.83 TC14 4.17 8.61 6.94 1.94 0.28 TC17 7.83 11.33 1.67 1.50 3.33 TC18 4.67 9.67 3.50 0.83 0.83 TC19 6.50 7.00 0.67 4.67 0.67 TC20 6.50 7.00 1.00 3.67 2.50 TC21 4.33 4.17 1.00 4.83 5.00 TC22 3.00 7.00 3.83 13.83 1.50 TC24 4.00 8.17 3.67 8.50 1.67 TC25 6.17 5.33 0 3.67 3.50 TC27 4.17 9.00 0.17 1.33 1.17 TC28 4.33 5.67 33.33 1.33 1.00 Mean ±sd 5.4±2.1 7.62±1.9 3.8±8.1 3.4±3.7 0.3±5 ' Epinephelus spp. and Mycteroperca spp. 499 Table 5. Twenty-five most frequently sighted fish species during roving diver surveys in the Turks and Caicos Islands, with mean densities for species recorded in belt transects. Scientific name Common name Sighting frequency’ (%) Density (#/100mh Thalassoma bifasciatum Bluehead Wrasse 97 - Halichoeres garnoti Yellowhead Wrasse 97 - Scarus taeniopterus Princess Parrotfish 97 1.32 Sparisoma viride Stoplight Parrotfish 94 2.08 Chromis cyanea Blue Chromis 91 - Stegastes partitus Bicolor Damselfish 88 - Gramma loreto Fairy Basslet 88 - Acanthurus coeruleus Blue Tang 88 3.65 Scarus croicensis Striped Parrotfish 88 1.84 Epinephelus fulvus Coney 88 1.27 Chaetodon capistratus Foureye Butterflyfish 88 1.38 Holocanthus tricolor Rock Beauty 85 0.59 Haemulon flavolineatum French Grunt 82 2.83 Caranx ruber Bar Jack 82 0.80 Acanthurus bahianus Ocean Surgeonfish 79 2.72 Melichthys niger Black Durgon 76 1.67 Ocyurus chrysurus Yellowtail Snapper 73 2.13 Sparisoma aurofrenatum Redband Parrotfish 70 1.32 Pseudupeneus maculatus Spotted Goatfish 67 - Bodianthus rufus Spanish Hogfish 67 0.33 Chaetodon striatus Banded Butterflyfish 67 0.43 Haemulon sciurus Bluestriped Grunt 64 1.53 Aulostomus maculatus Trumpetfish 64 - Stegastes planifrons Threespot Damselfish 61 - Scarus vetula Queen Parrotfish 61 1.89 ’Sighting frequency for the AGRRA surveys from the REEF database at http://www.reef org/data/twa/surveys/index.shtml 500 A Flower Garden Banks 95°50' 94°40' 93°30' 92'’2()' Figure 1. (A) AGRRA survey sites at the Flower Garden Banks. Location of (B) Buoy 5, West Flower Garden, (C) Buoy 2, East Flower Garden Bank. lEI=Buoy A RAPID ASSESSMENT OE THE FLOWER GARDEN BANKS NATIONAL MARINE SANCTUARY (STONY CORALS, ALGAE AND FISHES) BY CHRISTY V. PATTENGILL-SEMMENS* AND STEPHEN R. GITTINGS^ ABSTRACT Benthic and fish communities at one site on each of the East and West Flower Garden Banks were assessed using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol in August 1999. Surveys at 20-28 m revealed high coral cover (-50%) dominated by large (mean diameter 81-93 cm) healthy corals with total (recent + old) partial-colony mortality values averaging 13%. Turf algae were the dominant algal functional group and the mean relative abundance of macroalgae was <10%. The large abundance, size and biomass of many fishes reflected the low fishing pressure on the Banks. Due to their near-pristine condition, the Flower Garden Banks data will prove to be a valuable component in the rapid assessment database and its resulting determination of regional reef condition. INTRODUCTION The East and West Flower Garden Banks (EFG and WFG), located 175 km southeast of Galveston, Texas on the edge of the U.S. Gulf Coast continental shelf (Fig. lA), were created by the uplift of Jurassic-age salt domes. Rising about 100 m above the surrounding depths to within 18 m of the surface, the Flower Garden Banks (FGB) support the northernmost coral reefs in the continental United States. The low diversity (about 21 species), high cover and large size of stony corals, and the low abundance of benthic macroalgae relative to most Caribbean reefs have been well documented (Bright and Pequegnat, 1974; Bright et al., 1974; Boland et al., 1983; Dennis, 1985; Rezak et al., 1985; Dennis and Bright, 1988; Gittings et al., 1993). The FGB are also less susceptible to bleaching than most coral reefs because they are fairly deep; the 1998 regional mass bleaching event did not occur here. The FGB are dominated by massive boulder corals (particularly Montastraea spp. and Diploria spp.) but lack acroporids and shallow-water gorgonians. ' Reef Environmental Education Foundation and NOAA/National Marine Sanctuaries Program REEF, P.O. Box 246, Key Largo, FL 33037. Email: christy(@reef.org ^ NOAA/National Marine Sanctuaries Program, 1305 E. West, Hwy., SSMC-4, N/ORM62, #11645, Silver Spring, MD 20910 502 Fish diversity is also comparatively low (approximately 260 species) but abundances are high (Pattengill, 1998). Fish families and groups that are notably absent, or only represented by one or few species, include grunts (Haemulidae), snappers (Lutjanidae), and hamlets {Hypoplectrus sp.). The Banks are year-round habitat for manta rays {Manta birosths and Mobula hypostomd) and whale sharks {Rhincodon typus) and serve as a winter habitat for several species of schooling sharks including scalloped hammerheads {Sphyrna lewini), silky sharks {Carcharhinus falciformis), and spotted eagle rays {Aetobatus narinari) (Childs, 2001). A unique feature of the EFG is a brine seep at 72 m. The seep features a brine pool with a chemosynthetic bacterial assemblage that is known to be a significant exporter of carbon to the deeper parts of the Bank (Rezak et al., 1985). The seep also plays a significant role in the physiographic structure of the EFG due to the dissolution of salt which results in local faulting and subsidence. The EFG and WFG are managed and protected by the National Oceanic and Atmospheric Administration (NOAA)’s National Marine Sanctuary System and the Department of Interior’s Minerals Management Service. Together with Stetson Bank, they make up the Flower Garden Banks National Marine Sanctuary (FGBNMS). Anthropogenic impacts on the Flower Gardens are relatively low, mainly due to their distance from land. Very little fishing pressure exists on the reefs (see http://www.sanctuaries.nos.noaa.gov/oms/pdfs/FlowerGardensRegs.pdf; Subpart L or 15CFR922.122); spearfishing and techniques that disturb benthic habitats, including trawls, traps, and bottom long lines, are prohibited but hook-and-line fishing is permitted. The main, local source of human-induced disturbance is mechanical damage due to anchors, seismic cables, and occasionally long-line fishing tackle. Scuba diving is allowed on the Banks and moorings have been installed to reduce anchor damage. A long-term monitoring program has been in place for approximately 20 years. Historical biotic changes have been attributed primarily to regional or global events such as the die- off of Diadema antillarum and periodic coral bleaching. In August 1999, an AGRRA expedition to the FGB was coordinated by NOAA and the Reef Environmental Education Foundation (REEF) in conjunction with the annual FGBNMS REEF Field Survey for volunteer fish monitoring. The Banks were chosen for the survey because "end-member" reefs, including those that are unusually luxuriant, are particularly relevant to the AGRRA program (R. Ginsburg, personal communication). METHODS The survey team included seven scientists from the National Marine Sanctuary Program and three REEF experts. NOAA’s two long-term monitoring sites at EFG Buoy 2 and WFG Buoy 5 (Fig. IB, C) each of which is considered representative of the high- diversity stony coral zone on the Bank (Gittings et al., 1992), were strategically chosen for the surveys. The AGRRA protocol version 2.1 (see Appendix One, this volume) was used with the following modifications. Benthic and fish surveys were conducted simultaneously with divers dispersed over a radius of about 150 m around the moored 503 boat. Five divers conducted coral transects and algal quadrats during any given dive. Coral sizes were measured to the closest 5 cm, and any sediment deposited in the algal quadrats was gently removed by hand waving before estimating the relative abundance of crustose coralline algae. Cyanobacteria were counted as turf algae in 64 quadrats and as macroalgae in three quadrats at EFG. Three divers conducted the fish-belt transects and roving diver (RDT) surveys. Counts of serranids (groupers) were restricted to species of Epinephelus and Mycteroperca; scarids (parrotfishes) and haemulids (grunts) less than 5 cm in length were not tallied. All surveys were made between depths of 20 and 28m during daylight hours (7:00 a.m.-6:00 p.m.). The benthic- and fish-transect data were entered into a custom Excel spreadsheet provided by the AGRRA organizing committee. REEF provided the RDT data in the American Standard Code for Information Interchange (ASCII) format. The percent coral cover, percent mortality, mean colony size, incidence of disease and bleaching, and relative algal abundance were calculated and compared between Banks using a t-test. A macroalgal index (calculated as macroalgal index = % relative macroalgal abundance x canopy height-a proxy for macroalgal biomass) was also used as a comparison metric. Using the fish transects as replicates, the average density (#/100 m ) and size (cm) of each fish species and family recorded were calculated for each site. The average density and size of each species and family were compared between Banks using a t-test. Transect data were also used to calculate biomass for each fish species using standardized conversion equations (Appendix Two, this volume). The RDT survey data provided a species list, frequency of occurrence, and relative abundance data for each Bank. Percent sighting frequency (%SF) for each species was the percentage of dives in which the species was recorded. An estimate of abundance (den) was calculated as: Density score = ((nsxl)+(nFx2)+(nMx3)+(nAx4)) / (ns + np + nM + nA), where ns, np, nM, and nA represented the number of times each abundance category (single, few, many, abundant) was assigned for a given species. RESULTS Due to the minimum depth of the Banks, only one depth interval at each site was surveyed. At the WFG, 135 coral colonies and 55 algal quadrats were examined along 1 1 benthic transects and 1 1 RDT fish surveys and 12 fish-belt transects were performed. At the EFG, 160 coral colonies and 67 algal quadrats were surveyed on 14 transects and 15 RDT fish surveys and 12 fish-belt transects were conducted. Stony Corals Live stony coral cover averaged 54% and 49% at the WFG and EFG, respectively (Table 1). Nine species of “large” stony corals (with diameters >25 cm) were recorded within the transects at the WFG and 1 1 species at the EFG. Dominant species at the WFG (Fig. 2 A) were Montastraea franksi (40% of all colonies counted), Diploria strigosa (27%), Montastraea cavernosa (8%), and Montastraea faveolata (7%). The dominant 504 corals at the EFG (Fig. 2B) were Montastraea franksi (37%), Porites astreoides (16%), Montastraea cavernosa (13%), and Diploria strigosa (13%). The average diameter of the large corals (Table 2) was significantly greater (t-test; P<0.05) at the WFG than at the EFG (93 versus 81 cm, respectively). None of the transected corals showed any signs of disease. Parrotfish bites were reported in about 8% of all colonies surveyed. Pale bleaching was noted in some colonies (~6-16%). Overall, for colony surfaces recent mortality averaged 2%, mean old mortality was 1 1.5% and total partial mortality averaged 13%. The density of coral recruits found in the quadrats was 2.3 and 1 .7 per m^ at the WFG and EFG, respectively (Table 3). The recruits were P. astreoides and Agaricia agaricites, in equal proportions. A West Flower Garden (n=l 35):] 35) B East Flower Garden (n=160)i= 160) others * Montastraea franksi 40% Diploria strigosa 27% Porites astreoides 4% Colpophyllia natans 6% Montastraea faveolata 7% Montastraea cavernosa 8% Montastraea cavernosa 13% Montastraea franksi 37% Diploria strigosa 13% Porites astreoides 16% others* Colpophyllia 50/^ natans 6% Montastraea faveolata 9% * others include: Stephanocoenia intersepta, Madracis decactis and Agaricia agaricites *others include: Stephanocoenia intersepta, Madracis decactis, Montastraea annularis, Madracis mirabilis and Siderastrea siderea. Figure 2. Species composition and mean relative abundance of all stony corals (>25 cm diameter) at (A) West Flower Garden, (B) East Flower Garden Bank reefs. Algae Relative macroalgal abundance was very low (<10%) on both Banks (Table 3). Average macroalgal height was 1.0 cm, yielding macroalgal indices of 8.0 and 10.0 at the WFG and EFG, respectively. A mat cyanobacterium at the EFG that was common in the algal quadrats, on the sand flats, and on several coral heads was most likely responsible for the significantly higher (80% versus 72%) relative abundance values of turf algae at the EFG (t-test; p<0.05). Very few individuals of Diadema antillarum were sighted within the transects (0.9 and 1.4/100 m^ at the WFG and EFG, respectively). 505 Fish The average density of most families (Fig. 3) and species of fishes surveyed in the belt transects was similar at the EFG and WFG. Parrotfish were the most abundant fish recorded in the transects. The densities of graysby (Epinephelus cruentatus) and Spanish hogfish {Bodianus rufus) at the WFG were approximately twice those at the EFG (t-test; p<0.05). The density of reef butterflyfish {Chaetodon sedentarius) was also two and a half times greater at the WFG; however, this difference was not significant (t-test; p=0.059). Grunts, several species of parrotfish and snapper, hogfish {Lachnolaimus maximus), and gray angelfish {Pomacanthus arcuatus) were absent at both Banks, a distinguishing characteristic of the FGB’s fish assemblage (Pattengill, 1998). Figure 3. Mean fish density (no. individuals/100 ± sd) for eight fish families at the WFG (West Flower Garden) and EFG (East Flower Garden) Banks. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. A total of 1 1 7 fish species were seen by the AGRRA team during RDT surveys at the EFG and WFG. Great barracuda {Sphyraena barracuda), sharpnose puffer {Canthigaster rostrata), and black durgon {Melichthys niger) were documented in all surveys (Table 4). Species that were relatively common at the FGB compared to most other Caribbean reefs include longsnout butterflyfish {Chaetodon aculeatus), blue angelfish {Holacanthus bermudensis), and several species of jacks (Carangidae). Individuals of the golden phase of the smooth trunkfish {Lactophrys triqueter), a phase that is unique to the FGBNMS (Pattengill-Semmens, 1999), were also sighted. One new record for the Banks, a sharptail eel {Myrichthys breviceps), was recorded at the WFG. (An individual of the same eel species had been recorded on video earlier in the summer.) 506 Numerically, the most abundant species were: bluehead wrasse {Thalassoma bifasciatum), threespot damselfish (Stegastes planifrons), bicolor damselfish {S. partitus) queen parrotfish {Scarus vetula), along with planktivorous creolefish {Paranthias furcifer) and brown chromis {Chromis multilineata). Average sizes of parrotfishes and groupers were relatively high which resulted in relatively high biomass values (Table 5). The size frequency distributions of two feeding guilds, carnivores (select grouper genera and snappers) and herbivores (parrotfishes >5 cm, surgeonfish, and yellowtail damselfish, Microspathodon chrysurus), are shown in figure 4. Three-fourths of the individuals in the carnivore feeding guild were groupers (yellowmouth grouper, Mycteroperca interstitialis; tiger grouper, M. tigris; graysby, E. cruentatus; coney, E. fulvus; in descending order of density) with gray snapper {Lutjanus ghseus) making up the remainder. Approximately 45% of these carnivores were greater than 30 cm in length (Fig. 4 A) and the average size of the groupers was 25 cm. Approximately 70% of the individuals in the herbivore feeding guild were between 1 1 and 30 cm (Fig. 4B). The average size of the parrotfishes was 22 cm. 50 £ 40 •i 30 o Q. £ 20 a. 10 50 40 e 30 o Cl. o 20 10 A Carnivore Size Frequency □ WFG n=17 ■ EFG n=8 0-5 6-10 11-20 21-30 31-40 >40 Size (cm) B Herbivore Size Frequency □ WFG n= 153 ■ EFG n=162 [k 0-5 6-10 11-20 21-30 Size (cm) 31-40 >40 Figure 4, Size frequency distribution of (A) carnivores (lutjanids, select serranids) and (B) herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) at the WFG (West Flower Garden) and EFG (East Flower Garden) Banks. 507 For most species, average sizes recorded in the belt transects were similar between the EFG and WFG. Significant differences in length were detected by a t-test in the blue tang {Acanthurus coeruleus), which was longer at the WFG, whereas princess parrotfish (Scarus taeniopterus) and yellowtail damselfish were longer at the EFG. DISCUSSION Results from this assessment revealed reefs with high stony coral cover that are dominated by large boulder corals. Pale bleaching was evident in some surveyed colonies but very little disease was noted and only on stony corals outside the benthic transects. Macroalgal biomass was very low. Results of the fish surveys showed an assemblage that is relatively low in diversity but high in biomass. The Banks appeared to support fewer, but larger, individual fishes in comparison to other Caribbean-area reefs. The large average size of parrotfishes and groupers likely reflected the low fishing pressure on the reefs. Recruitment success appeared to be driving the size differences for three of the surveyed fishes, giving a high abundance of juvenile blue tang on the EFG and a high number of juvenile princess parrotfish and yellowtail damselfish on the WFG. An additional 15 REEF volunteers conducted 74 RDT surveys during the cruise. These data were not included in the AGRRA data set but were added to the REEF database, which can be accessed from the REEF Website (http://www.reef org). As a result of annual field surveys at the Banks between 1993 and 1999, a total of 1,495 REEF surveys have been generated for the FGBNMS (over 1,100 survey hours). These data represent a valuable source of information for the Sanctuary management. To date, 257 fish species have been documented at the FGBNMS. A comprehensive fish species list for the FGB has been published using these data (Pattengill, 1998). The FGB are deep reefs on offshore banks that are far removed from land and experience little anthropogenic disturbance. Described as “near pristine,” they provide an important piece of the regional picture of Mesoamerican reef condition. The reefs of the FGB have been, and continue to be, well studied. A long-term monitoring project has been in place for over 20 years. Data collected during the AGRRA assessment corroborated findings of previous studies on the condition of the FGB ecosystem. The importance of these data will be further highlighted when comparable data from dozens of sites are compiled to create a more complete picture of the current status of western Atlantic coral reefs. The FGB can then be used to help “calibrate” the AGRRA “scale” of reef condition. ACKNOWLEDGMENTS This project was made possible by the AGRRA research team: S. Bernhardt (Texas A&M University), K. Deaver (REEF), S. Gittings (NOAA), D. Mizell (REEF), C. Ostrom (NOAA), C. Pattengill-Semmens (REEF), G.P. Schmahl (NOAA), T. Shyka (NOAA), P. Souik (NOAA), and M. Tartt (NOAA). The support of the M/V Spree crew and Rinn Boats, Inc. is appreciated. Funding support for this cruise was provided by the 508 Flower Garden Banks National Marine Sanctuary, the Science Team of the Marine Sanctuary Program, the Reef Environmental Education Foundation and its volunteers. The surveys provided by the REEF volunteers are also greatly appreciated. REFERENCES Boland, G.S., B.J. Gallaway, J.S. Baker, and G.S. Lewbel 1 983. Ecological effects of energy development on reef fish of the Flower Garden Banks. Final Report. LGL Ecological Research Associates. Elouston, TX. 466 pp. Bright, T.J., and L.H. Pequegnat 1974. Biota of the West Flower Garden Bank. Gulf Publishing Co., Houston, TX. 435 pp. Bright, T.J., W.E. Pequegnat, R. Dubois, and D.A. Gettleson 1974. Baseline Survey Stetson Bank Gulf of Mexico. Signal Oil and Gas Co., Texas A&M University, Dept, of Oceanography, College Station, TX. 38 pp. Childs, J. 2001 . The occurrence, habitat use, and behavior of sharks and rays associating with topographic highs in the northwestern Gulf of Mexico. M.S. Thesis, Texas A&M University, College Station, TX. 213 pp. Dennis, G.D. 1 985. Tropical reef fish assemblages in the northwestern Gulf of Mexico. M.S. Thesis, Texas A&M University, College Station, TX. 184 pp. Dennis, G.D., and T.J. Bright 1988. Reef fish assemblages on hard banks in the northwestern Gulf of Mexico. Bulletin of Marine Science 43:280-307. Gittings, S.R., K.J.P. Deslarzes, D.K. Hagman, and G.S. Boland 1992. Reef coral populations and growth on the Flower Garden Banks, northwest Gulf of Mexico. Proceedings Seventh International Coral Reef Symposium 1:90-96. Gittings, S.R., T.J. Bright, and D.K. Hagman 1993. Protection and monitoring of reefs on the Flower Garden Banks, 1972-1992. Pp. 181-187. In: R.N. Ginsburg (compiler). Global Aspects of Coral Reefs - Health, Hazards, and History. Rosenstiel School of Marine and Atmospheric Science, University of Miami, Miami. Pattengill, C.V. 1 998. The structure and persistence of reef fish assemblages of the Flower Garden Banks National Marine Sanctuary. Ph.D. Dissertation, Texas A&M University, College Station, TX. 176 pp. Pattengill-Semmens, C.V. 1 999. Occurrence of a unique color morph in the smooth trunkfish (Lactophrys triqueter) at the Flower Garden Banks and Stetson Bank, northwest Gulf of Mexico. Bulletin of Marine Science 65:587-591. Rezak, R., T.J. Bright, and D.W. McGrail 1985. Reefs and Banks of the Northwestern Gulf of Mexico: Their Geological, Biological and Physical Dynamics. John Wiley and Sons, New York. 259 pp. 509 C/5 I 5 cm) (>5 cm) WFG#5 960.4 ± 864.5 4263.1 ±2984.0 0 376.9 ± 736.5 554.3 ± 1041.3 EFG #2 493.2 ± 370.3 4765.2 ± 3387.6 0 252.8 ± 875.6 343.4 ±473.0 ' Epinephelus spp. and Mycteroperca spp. 511 Plate 12 A. The formerly ubiquitous, herbivorous sea urchin, Diadema antillarum, played a key role in preventing overgrowth of stony corals by macroalgae in many areas of the wider Caribbean prior to its regionwide demise in 1983-1984. Localized population increases are currently being reported, although densities everywhere are still far below pre-dieoff levels. (Photo Andrew W. Bruckner) Plate 12B. On many Caribbean reefs lacking large sized herbivorous fishes, the loss of D. antillarum has allowed macroalgae to colonize coral skeletons and then overgrow the living coral tissues, as shown for the Lobophora variegata on this Colpophyllia natans. The result has been a shift to algal dominated reefs in some areas, as well as increased mortality and reduced recruitment of stony corals. (Photo Andrew W. Bruckner) 67 05' 66«’55’ 66°45’ 66‘''’35’ 512 Figure 1. AGRRA survey sites in the Archipielago de Los Roques National Park, Venezuela. RAPID ASSESSMENT OF CORAL REEFS IN THE ARCHIPIELAGO DE LOS ROQUES NATIONAL PARK, VENEZUELA (PART 1: STONY CORALS AND ALGAE) BY ESTRELLA VILLAMIZAR,*’^ JUAN M. POSADA, and SANTIAGO GOMEZ'^ ABSTRACT The status of the coral reefs in Archipielago de Los Roques National Park, Venezuela was appraised in October, 1999 at 13 sites using the Atlantic and Gulf Rapid Reef Assessment (AGRRA) protocol. The reef complex appears to be in good condition overall. Live stony coral cover was especially high (30-60%) at four sites between 8-13 m in the southern and southwestern reefs. However, “standing dead” Acropora were particularly conspicuous in the eastern barrier reef and yellow-band disease was present in some colonies of Montastraea annularis and M. faveolata. Average values of recent partial-colony mortality were 8-10% in stony corals of >10 cm diameter in three reefs. Macroalgae were scarce (<10% relative abundance) in most (12/13) sites. The Los Roques reefs are likely to serve as important sources of coral larvae traveling through the southern Caribbean. Their protected status must be maintained. INTRODUCTION The Archipielago de Los Roques National Park is Venezuela’s major coral reef complex and one of the most important in the southern Caribbean region. Located 150 km north of the central Venezuelan coast, it is outside the geographic area routinely influenced by hurricanes. Land-based anthropogenic impacts are reduced in comparison to those in Venezuela’s continental reef systems. The park, however, is subject to the impacts of commercial fishing pressures (mostly for spiny lobsters) and to increasing touristic activity (see Posada et ah, this volume). Regional diseases, in particular white- * Institute de Zoologia Tropical, Universidad Central de Venezuela, Apartado 47058, Caracas 1041-A, Venezuela. Email; evillami@strix.ciens.ucv.ve 2 Fundacion Cientifica Los Roques, Apartado 1 139, Caracas 1010-A, Venezuela. ^ Departamento de Biologia de Organismos, Universidad Simon Bolivar, Apartado 89000, Caracas 1080-A, Venezuela. Email; jposada@usb.ve Centro de Botanica Tropical, Universidad Central de Venezuela, Apartado 47114, Caracas 1041-A, Venezuela. Email; sagomez@strix.ciens.ucv.ve 514 band disease (WBD), have also impacted its stony coral populations, particularly the important framework hmXdQV Acropora palmata (Garcia, 2001; Leon, 2001), The archipelago, which developed during the late Pleistocene around a metamorphic-igneous platform centered in Gran Roque Island, includes two main barrier reefs. The 24 km long eastern barrier, which runs from northeast to southeast, is fully exposed to the prevailing winds and seas. Strong currents characterize its seaward side. The southern barrier extends about 30 km from southeast to southwest and is progressively more protected from waves and currents as one moves from east to west. By protecting the interior of the archipelago from the direct influence of oceanic waves and currents, the barriers have facilitated the formation of numerous patch reefs, sand cays with fringing reefs, seagrass meadows and, in the southeastern part of the archipelago, mangrove forests (Mendez-Baamonde, 1978). Annual surface seawater temperatures at Los Roques average 26° C (sd=2° C) with a minimum of about 23° C in February and a maximum of about 30° C in May although temperatures as high as the maximum can be found during October and November (Villamizar, 1993). Vertical water transparency is relatively high (15 to 25 m). Mendez-Baamonde (1978) provided the first qualitative description of the major reef habitats in the archipelago. However, much of the research which has been undertaken in the Los Roques coral reefs is either unpublished or in undergraduate theses. Hung (1985) found that species distribution patterns, numbers of colonies and live stony coral cover varied with habitat in the southwestern Dos Mosquises Sur (DMS) fringing reef Coral cover was greatest (7,200 cm^/m^) at “intermediate depths” (3-18 m) in the terrace and slope zones which were dominated by the Montastraea annularis complex. Hung (1985) also found numerous dead colonies of Acropora palmata in the reef crest at DMS and at another nearby fringing reef [Dos Mosquises Herradura (DMH)]. Sandia and Medina (1987), who studied the population dynamics of A. cervicornis off DMS, reported a maximum dominance (-18.0% of live coral cover and 9% of the colonies) of this species on the fore-reef terrace in depths of 3-6 m where numerous dead fragments of A. cervicornis were also found. A. palmata was the dominant species in shallower (<3 m) water while massive corals (particularly M. annularis and Colpophyllia natans) and gorgonians were most common below 6 m. Subsequently, Crdquer and Villamizar (1998) studied the effect of topographic variation on reef community structure in four sites at DMS finding the largest number of stony coral species (24) in the site with the least slope. Nevertheless the between-site similarity index of species composition was higher for the reef-slope zone than for the reef flat, reef crest and deep reef flat. Garcia (2001) found that 6.87 % of the stony corals (n=3,344) examined during 2000 in seven Los Roques localities were affected by disease. Yellow-band (yellow- blotch) disease (YBD) and dark spots disease (DSD) were the most common (both 2.1%) with other diseases each occurring in fewer than 1% of the colonies. Le6n (2001) has assessed the effects of recreational diving activities by comparing one each experimental (frequently visited by divers) and control (low touristic activity) site in several of the Los Roques reefs. Live stony coral cover overall averaged 32.6 % (sd=16) and was higher in the control sites than in the experimental sites in three reefs [including Crasqui-La Venada (CRV) with 49% versus 1 1% and Pelona de Rabusqui (PR) with 53% versus 16%, respectively] and lower in the controls only in the 515 DMS reef. Leon (2001) also found that YBD and DSD were the most common stony coral diseases at these sites in 2000. The purpose of the present study was to evaluate the condition in October 1 999 of the Los Roques coral reefs through the application of the AGRRA protocols. Stony coral and algal components are considered in this paper; the fish assessment is given by Posada et at. (this volume). Our results provide a comprehensive overview of the archipelago’s reefs, allowing comparisons with other areas in the Caribbean (Kramer, this volume), and creating a baseline against which future assessments can be compared. METHODS Thirteen sites which are considered by the two senior authors to be representative of the barrier, fringing, and patch-reef habitats at Los Roques were chosen for assessment. The majority of the sites were located along the axes of the two main barrier reefs in recognition of their size and ecological importance; only the northwest sector of the archipelago was not surveyed. Three divers employed the AGRRA Version 2.2 benthos protocol (see Appendix One, this volume) with the following modifications. The minimum diameter of individually surveyed corals was 10 cm. Coral size was recorded to the nearest 5 cm. Species of stony corals that remain small as adults (e.g., Favia fragum) were omitted from the counts of coral recruits. Sediments were removed fairly vigorously from the algal quadrats before estimating the abundanee of crustose coralline algae. Training in stony coral identifications and in designation of the percentage of partial mortality and algal abundance was condueted the day before the surveys officially began. Field identifications of corals were corroborated by reference to Humann (1994). Correlation coefficients were calculated to examine relationships between the percentage of standing dead corals and both the abundance and partial mortality of A. palmata. Distributions were normalized by using the log N+1 transformation for abundances and the arcsine root transformation for proportions. RESULTS The 1 3 surveys were conducted at four sites each in the eastern and southern barrier reefs, at two sites each in the central-eastern and southwestern fringing reefs, and in one northeastern patch reef (Table 1, Fig. 1). Back-reef habitats were surveyed in the eastern barrier [Cayo Vapor (CV), Boca del Medio (BM), Barrera Este (BE) and Boca de Sebastopol (BS)] where our access to the fore reef was restricted by high waves. One outer reef crest [Boca Cote Somero (BCS)] and three fore reefs [Punta Cayo Sal (PCS), Cayo Sal Sur (CSS) and Boca del Cote Profundo (BCP)] were visited in the southern barrier, two of which (BCS, BCP) were loeated near a large mangrove forest. The reef at DMS has a maximum depth of 40 m, but the remaining fringing reefs (PR, CRV and DMH) extend no deeper than 6.5, 10, and 12 m, respectively, and the base of the Noronqui de Abajo (NOR) patch reef is at 2 m. Eight of the survey sites were considered shallow (<6.5 m) and five as being in deeper water (7.5-13.5 m). 516 Stony Corals The total cover of live stony corals (Table 1) varied from <10% in four shallow reefs (BS, CV, CRV, BCS) to >50% in three of the deeper southern reefs (PCS, BCP and DMS). Live coral cover in the remaining sites ranged between about 20 and 30%. The density of stony corals that were at least 1 0 cm in diameter was higher overall in the southern sites than along the eastern side of the archipelago (Table 1). A total of 27 scleractinians (including Eusmilia fastigiata, Favia fragum and Porites branneri in addition to species on the original AGRRA coral list) and one hydrozoan (Millepora complanata) were large enough (>10 cm in diameter) for individual survey. Seven to 13 of these species occurred in each of the shallow reefs where colonies of Acropora palmata that were mostly dead (see below) numerically dominated most (six/eight) of the reefs, along with Porites astreoides and Diploria strigosa (Fig. 2A). Live A. cervicornis was found in only two of the shallow reefs (NOR, where A. palmata was rare, and DMH). M annularis occurred in six of the shallow reefs (BCS, DMH, CV, PR, CRV and NOR). Although M. faveolata was present in seven shallow reefs, it was relatively abundant only in the deepest (PR). M complanata was abundant in one shallow reef on the southern barrier (BCS). Eleven to 17 species of the individually surveyed corals were recorded in each of the deeper reefs. The most common species, M. faveolata, M. annularis, M. cavernosa A Shallow Sites B Deeper sites Porites porites 3% Other 7% Porites astreoides 16% Millepora complanata 7% Montastraea faveolata Acropora cervicornis 6% Agaricia 8% Montastraea Diploria annularis strigosa 7% 10% Other Stephanocoenia inter septa 4% Porites porites 4% Porites Millepora complanata4% Montastraea cavernosa 12% Acropora palmata\0% Agaricia agaricites4% Colpophyllia natanslVo Diploria labyrinthiform is 4% Montastraea annularis 8% Montastraea faveolata 22% Figure 2. Species composition and mean relative abundance of the most abundant stony corals (>10 cm diameter) in (A) shallow (n=672) and (B) deeper (n=415) sites in Los Roques, Venezuela. Other: (A) = Agaricia tenuifolia, Colpophyllia natans, Diploria clivosa, D. labyrinthiformis, Favia fragum, Porites branneri, Siderastrea radians, S. siderea; (B) = A. tenuifolia, Dendrogyra cylindrus, Dichocoenia stokesi, D. clivosa, D. labyrinthiformis, Eusmilia fastigiata, F. fragum, Madracis decactis, M. mirabilis, Meandrina meandrites, Montastraea franksi, P. branneri, S. radians, S. siderea. 517 (especially at CSS) and C. natans (in particular at DMS), all had massive morphologies (Fig. 2B) except in the eastern barrier reef (at BM) where A. palmata, P. astreoides and Agaricia agaricites were predominant. Average values in excess of 100 cm (Table 2) were found for the maximum diameters of individually surveyed stony corals in the five eastern reefs that were numerically dominated by A. palmata (BE, BS, CV, CRV and BM). The smallest (38 cm) was in one of the deeper, southern barrier reefs (CSS) where M cavernosa was the most abundant species and the largest colonies of M. faveolata were less than 120 cm in maximum diameter. Average height was closely correlated with maximum diameter in most sites (Fig. 3). 300 250 200 B E u JS •Sf *s JS a « 180 160 140 - 120 - 100 80 60 40 - 20 - 0 Shallow (N = 672) oc w (JO > > C/D o CQ CQ u Di a. u Z u CQ Sites ShaUow (N = 672) oi w CZ) > > oi C/3 o CQ CQ U 0^ Oh u 2 u CQ Sites Deeper (N = 415) >5 Dh OO C>0 U c/2 U CQ PQ U Oh Deeper (N = 415) Figure 3. (A) Diameter and (B) height (mean ± sd) of all stony corals (>10 cm diameter) by site in Los Roques, Venezuela. See Table 1 for site codes. 518 Many colonies of^. palmata in the shallow reefs were >100 cm in maximum diameter and the largest exceeded 400 cm (Fig. 4). A majority of the D. strigosa were <100 cm although some were larger (to 210 cm). P. astreoides was almost exclusively represented by colonies with diameters of <20 cm. M faveolata in the deeper reefs resembled the shallow A. palmata in having a wide range of maximum diameters with the largest colonies being ~300 cm in maximum diameter (Fig. 5). Most colonies of the smaller-sized M. cavernosa were in the 20-70 cm size intervals. Vi .Si ‘e o "o U % Ov " m 0^ Diameter (cm) Diameter (cm) Figure 4. Size-frequency distributions of >10 cm diameter colonies of Acropora palmata, Diploria strigosa and Porites astreoides at shallow sites in Los Roques, Venezuela. No stony coral “recruits” (^2 cm diameter) were recorded in seven (three shallow, four deeper) of the survey sites (Table 3). Their densities in the remainder were less than five corals/m (~0.3/.0625 m ) at all but two of the shallower sites (BS, PR). Agaricia and Porites were the most common genera. Agaricia recruits appeared to preferentially colonize dead A. palmata branches (particularly in BM). Although no recruits of Acropora, Montastraea or other large massive corals were found in the quadrats, some small established A. cervicornis recruits were noted outside the AGRRA survey site in the crest of the patch reef (NOR). 519 M faveolata N = 94 Diameter (cm) CD O ■*- CSJ Diameter (cm) Figure 5. Size-frequency distributions of >10 cm diameter colonies of Montastraea faveolata and M. cavernosa at deeper sites in Los Roques, Venezuela. Stony Coral Condition No totally bleached corals were observed at the time of our surveys. At seven sites, however, fewer than 4% of the colonies were either pale or partially bleached (Table 2). There was no evidence of infection with WBD in any of the live colonies of Acropora. YBD was found in three sites (Table 2), occurring in Montastraea annularis (seven colonies), M. faveolata (three colonies), and M franksi (one colony). A total of one colony each of Porites porites, P. astreoides and A. palmata also showed signs of a disease resembling YBD in overall appearance. One colony of M faveolata had signs resembling those of red-band disease and a Diploria strigosa had white plague. 520 Average values of recent partial-colony mortality (hereafter recent mortality) varied from <1% in two of the deeper reefs in the southern barrier to >8% in three shallow reefs (CRV, PR and BCS) (Table 2). In addition to the above-mentioned diseases, mortality agents identified in the shallow localities included predation by scarids and the snail Coralliophila plus overgrowth by gorgonians and algae. However, a large number of the shallow colonies with recent mortality were associated with territories of Stegastes planifrons (see below). Recent mortality was somewhat lower overall in the deeper sites (Table 2, Figs. 6, 7). We observed no predation by Diadema antillarum on live stony corals, although such behavior had been previously observed (Villamizar, personal observations during 1995 in DMH). Old partial mortality (%) _ — I — I 1 1 1 1 1 30 40 50 60 70 80 90 100 Recent partial mortality (%) Figure 6, Frequency distributions of (A) old partial-colony mortality and (B) recent partial- colony mortality of all stony corals (>10 cm diameter) at shallow sites in Los Roques, Venezuela. 521 Recent partial mortality (% ) Figure 7. Frequency distributions of (A) old partial-colony mortality and (B) recent partial-colony mortality of all stony corals (>10 cm diameter) at deep sites in Los Roques, Venezuela. Old partial-colony mortality (hereafter old mortality) ranged from <30% in six sites (two shallow, four deeper) to >60% in three of the shallow sites (Table 2). Total (recent + old) partial mortality was <25% in three of the four deeper southem-to- south western sites, -43-75% in the four reefs along the eastern barrier, and over 80% in one of the central-eastern fringing reefs (CRV). Of the colonies with some remaining live tissues (i.e., for which total mortality was <100%), recent mortality values for the dominant species varied from about 1% for M cavernosa and C. natans in the deeper sites to approximately 18% for A. palmata in shallow sites (Fig. 8). The eorresponding averages for old mortality were lowest in shallow^, cervicornis (-13%) and highest in D. strigosa (38%). The ratio of recent mortality to old mortality was lowest in the deeper M cavernosa (0.04) and highest in the sfidiWosN Acropora spp. (0.62 in^l. palmata; 0.64 m.A. cervicornis). 522 Figure 8. Percent old partial-colony mortality and percent recent partial colony mortality of >10 cm diameter stony corals with at least some live tissues (i.e., for which mortality is <100%) at shallow and deep sites in Los Roques, Venezuela. AP = Acropora palmata, AC =.A. cervicornis, DS = Diploria strigosa, PA = Porites astreoides, MAP = Montastraea faveolata, MC = M cavernosa, CN = ColoonhvUia natans. Over 25% of the colonies were “standing dead” (completely dead but still attached to the substratum) in six survey sites. All but one of these, BM, were shallow and five sites were located in (BS, CV, BE, BM) or near (CRV) the eastern barrier. Acropora palmata accounted for most of these standing dead with <7% (17/252) of all surveyed colonies having any live tissues. Statistically significant correlations were found between the percentage of standing dead corals in each site and both the abundance of A. palmata (r=0.93, p<0.05, n=13) and the percentage of its total mortality (r=0.87, p<0.05, n=13). Only one reef (DMH) had a high proportion (1 1/14) of colonies of A. palmata still having some living tissues. Elsewhere, with the exception of one or two colonies in each site that were partially alive, the remainder were standing dead. Algal groups, Diadema and Damselfishes Overall, turf algae clearly dominated the algal functional groups in seven shallow and three deeper sites (Table 3). Crustose coralline algae (especially Lithophyllum) and Peyssonnelia were dominant in two deeper reefs (CSS, DMS) and one shallow (BCS) reef and both were essentially codominant with turfs in one shallow fringing reef (CRV). Macroalgae were completely absent from quadrats in three sites (one shallow, two deeper) and conspicuous only at Boca de Cote shallow (BCS) where their relative abundance was 29% at 2.5 m. Commonly observed macroalgae were Halimeda opuntia, H. monile, Bryopsis pennata, Cladophoropsis sp., Ventricaria sp., Wrangelia penicillata. 523 Ceramium rubrum, Amphiroa fragilissima, and Dyctyopteris delicatula. The cyanobacterium, Blennothrix lyngbyacea, was included in the macroalgal group. Macroalgal indices (macroalgal relative abundance x macroalgal height) were 0-1 at nine sites, 2-10 at three sites, but 70 in one shallow site (BCS, where macroalgae also grew somewhat taller than the average elsewhere). The sea urchin, Diadema antillarum, was present in five shallow sites (Table 3). Maximum densities of about 36 individuals/ 100 m were found in a shallow fringing reef (DMH). Damselfish, particularly Stegastes planifrons, exhibiting algal-gardening behavior in individually surveyed stony corals were recorded in all but one site (CSS), with the highest percentages found in shallow water (Table 2). Algal gardens were primarily found in live colonies of M. faveolata and were also present in M. annularis, M. franksi and Millepora complanata; in one site (CRV), the damselfish showed a high preference for standing dead colonies of A. palmata. DISCUSSION The application of the AGRRA protocol, in concert with the recent studies of Garcia (2001) and Le6n (2001), has given us a general overview of the present condition of the barrier, fringing, and patch reefs in Los Roques National Park. The earlier descriptions of Mendez-Baamonde (1978) can now be updated and our results will allow comparisons of this important oceanic marine system with other reefs in the wider Caribbean (Kramer, this volume). Hung’s (1985) area-based estimate of percent live stony coral cover in a shallow SW fringing reef (DMH) in 1984 was highest at 2-3 m (mean=59, sd=17.5) which is about double that of our linear cover estimate in the same habitat in 1999 (Table 1). Numerous taxa {Acropora prolifera, Agaricia sp., Dendrogyra cylindrus, Madracis spp., Meandrina meandrites, Agaricia lamarcki) that routinely achieve diameters in excess of 10 cm were listed by Hung (1985) but absent from the AGRRA surveys. Although the A. prolifera may well have disappeared as a result of WBD, we are unsure if the apparent loss of the other taxa is real or an artifact of different sampling intensity. Although we found no diseased corals in this site, Garcia (2001) reported that 5.8% of the 605 stony corals (all size classes) examined here in 2000 showed signs of disease. In 1984, the percent of live stony coral cover in the DMS fringing reef was highest (mean=72, sd=22.5, areal estimate) at 3-18 m (Hung, 1985). Relative to that baseline, our linear-based percentage data at 12 m in 1999 (mean=60, sd=18.5) and Le6n’s (2001) areal estimate at 1-11 m in the experimental diving area (mean=52.8, sd=25.2) are somewhat reduced, but in Leon’s (2001) control area live stony coral coverage was substantially lower (mean=33.5, sd=3.2). Of the seven sites studied in 2000 by Garcia (2001), the frequency of diseased corals (11.6%, n=335 corals) was greatest in the DMS fringing reef and an order of magnitude larger than what we had found the previous year (Table 2). Garcia (2001) also noted proportionately higher incidences of disease in 2000 in two other sites that were visited by the AGRRA team (6.7% at CRV; n=492 corals, and 5.8% at Boca de Cote; n=542 corals). However, the two datasets are not strictly 524 comparable since Garcia (2001) examined all size classes and surveyed in slightly different depths (1.5-5 m and 4-9 m). Overall, during the quarter century since Mendez-Baamond’s (1978) pioneering observations, the most striking ehange in shallow-intermediate depths at Los Roques has been the near disappearance of species of Acropora. Many colonies died during the early 1980’s (Hung, 1985; Medina and Sandia, 1987), presumably from the effects of WBD as happened in so many areas of the wider Caribbean (Gladfelter, 1982; Aronson and Precht, 2001). Live A. palmata were still very rare in 1999 at the AGRRA survey sites with over half of the surviving colonies (1 1/17) occurring in a single shallow reef (DMH). Live colonies of A. cervicornis in the shallow sites were somewhat more abundant (e.g., 19 at DMH, 1 1 at NOR), more widespread (also occurring outside the belt transects in CRV, PR and BCS) and had a higher ratio of live/dead colonies (35/42). As we saw no recruits of A. palmata and those of A. cervicornis were only present in one site (NOR), it is very unlikely that their populations ean recover any time in the near future to levels eomparable to those observed in the archipelago during the 1970s. In contrast to the reduced coverage in many shallow reef habitats, live stony coral cover values of -30-60% at 8-13.5 m in the southern and southwestern reefs are high relative to recent reports from comparable habitats throughout much of the wider Caribbean (Kjerve, 1998; Kramer, this volume). The high proportion of large colonies of A. palmata and D. strigosa in shallow sites (Fig. 4) and of M. faveolata in deeper sites (Fig.5) is evidence that, prior to the arrival of WBD, conditions for the growth and survival of stony corals at Los Roques must have been excellent for many decades or even centuries. Smaller colonies of these species may reflect past recruitment (although recruits of these species were absent in the AGRRA surveys) or subdivision of colonies from physical breakage or partial mortality of the live tissues. The most common and widespread disease recorded in the 1 999 AGRRA assessment was YBD (Table 2), which primarily occurs in Montastraea, and it was tied for first place with DSD in Garcia’s (2001) surveys during 2000. Given the seminal importance of Montastraea spp. (Figs. 2, 5) in these habitats, it would be of great concern should this trend continue. In all but two shallow localities (BM, DMH), the key herbivore, Diadema antillarum, was either very rare or absent. Macroalgae were extremely sparse (<10% relative abundance) in all reefs except Boca de Cote, where the abundance of herbivorous fishes was somewhat reduced in the deeper (BCP) site (Posada et al., this volume) and where macroalgal indices (a proxy for biomass) were only elevated in the shallow (BCS) site. Hence, we suspeet that nutrients are naturally exported from the large mangrove forests nearby in Sebastopol and stimulate benthic macroalgal production at Boca de Cote, particularly in shallow water where proportionately more sunlight is available for photosynthesis. From the results obtained in this study, it is possible to conclude that the Archipielago de Los Roques as a whole is still a “healthy” coral reef system, even allowing for the drastic declines that have occurred in its populations of^. palmata and A. cervicornis. However, the deeper sites in the southern barrier and a southwestern fringing reef were in better condition than the shallow sites. We suggest that major efforts 525 must be made to continue to protect this reef system, which is of great importance for Venezuela and the Caribbean region. ACKNOWLEDGMENTS Special thanks to Patricia Kramer for helping us with the initial training in the application of AGRRA methods and with the surveys at Los Roques. Thanks to the Caribbean Environment Programme of the United Nations Environment Programme for its support of this study. The AGRRA Organizing Committee facilitated financial support as described in the Forward to this volume. We are grateful to fNPARQUES for permission to conduct the surveys at Los Roques and to Fundacion Cientifica Los Roques for providing access to their research facilities at Dos Mosquises. Judith Lang and Robert Ginsburg provided constructive comments and helpful advice during the preparation of the manuscript. REFERENCES Aronson, R.B., and W.F. Precht 2001. Evolutionary paleoecology of Caribbean coral reefs. Pp. 1771-233. In: W.D. Allmon and D.J. Bottjer (eds.). Evolutionary Paleoecology: The Ecological Context of Macroevolutionary Change. Columbia University Press, New York. Croquer, A., and E. Villamizar 1 998. Las variaciones de la pendiente topografica, un factor a considerar en la evaluacion de la estructura de una comunidad arrecifal. Revista de Biologla Tropical 46 Suplemento 5:29-40. Garcia, A. 2001 . Enfermedades y otras Anomallas que Afectan a los Corales Escleractlnidos en algunas Localidades del Parque Nacional Archipielago de Los Roques. Tesis de Grado, Escuela de Biologia, Universidad Central de Venezuela, Caracas, Venezuela, 169 pp. Gladfelter, W.B. 1982. White band disease in A. palmata: implications for the structure and growth of shallow reefs. Bulletin of Marine Science 32:639-643. Humann, P. 1994. Reef Coral Identification. Florida, Caribbean Bahamas. New World Publications, Inc. Jacksonville, Florida, 239 pp. Hung, M. 1985. Los Corales Petreos del Parque Nacional Archipielago de Los Roques. Tesis de Grado, Escuela de Biologia, Universidad Central de Venezuela, Caracas, Venezuela. 204 pp. Kjerfve, B. (editor) 1998. CARICOMP-Caribbean coral reef seagrass and mangrove sites. Coastal region and small island papers 3, UNESCO, Paris, France. 345 pp. 526 Leon, A. 2001 . Evaluacidn del Efecto de las Actividades de Buceo Recreativo en la Estructura Comunitaria de algunos Arrecifes Coralinos del Parque Nacional Archipielago de Los Roques. Tesis de Grado, Escuela de Biologia, Universidad Central de Venezuela, Caracas, Venezuela. 121 pp. Mendez-Baamonde, J. 1978. Archipielago Los Roques/Islas de Aves. Cuademos Lagoven. Caracas, Venezuela. 48 pp. Ramirez, V.P. 2001. Corales de Venezuela. Universidad de Oriente, Porlamar, Venezuela. 219 pp. Sandia, J., and R. Medina 1 987. Aspectos de la Dindmica Poblacional de Acropora cervicornis en el Parque Nacional Archipielago de Los Roques. Tesis de Grado, Escuela de Biologia, Universidad Central de Venezuela, Caracas, Venezuela. 123 pp. Villamizar, E. 1993. Evaluacidn de la Comunidad de Peces en Praderas de Fanerogamas marinas del Parque Nacional Archipielago de Los Roques. Tesis de Doctorado, Postgrado En Biologia, Mencion Ecologia, Facultad de Ciencias, Universidad Central de Venezuela, Caracas, 244 pp. Table 1. Site information for AGRRA stony coral and algal surveys in the Archipielago de Los Roques National Park, Venezuela. 527 Table 2. Size and condition (mean ± standard deviation) of all stony corals (>10 cm diameter) by site in Los Roques National Park, Venezuela. 528 = pale; PB = partly bleached ^WP = white plague; RB = red-band disease; YBP like = unknown disease resembling YBD; YBP = yellow-blotch disease; UK = cause of recent mortality unknown but disease suspected. 529 5 cm and acanthurids) was higher in the shallower (<8 m) reefs than at depths of 8-12 m; commercially important carnivores (lutjanids, select serranids) were relatively scarce in the shallow eastern sites. There was a significant inverse relationship between total fish density and live stony coral cover. The 11-20 cm size class dominated the length frequency distributions of the key herbivores and carnivores. The Los Roques fish community appears to be in good condition overall and least disturbed anthropogenically in the southern barrier reef INTRODUCTION The Archipielago de Los Roques National Park, established in 1 972, is one of the oldest national marine parks in the Caribbean and is Venezuela’s most important insular reef complex. Located 150 km north of the central Venezuelan coast, from 1 1° 44'-l 1° 58'N latitude and from 66°32-66°57'W longitude (Fig. 1), it covers an area of approximately 800 km . The 42 cays and 200 sand banks collectively form an irregular oval which is delimited to the east and south by narrow barrier reefs (approximately 20 and 30 km long, respectively) that partially enclose a shallow lagoon with a mean depth of 4 m. The lagoonal floor is predominantly bare sand or mud with important expanses of seagrass and/or macroalgal beds and numerous patch reefs. The archipelago’s location makes it especially favorable for reef development for three reasons: 1) physical environmental stability is high because it is not subject to the direct impact of hurricanes; 2) anthropogenic disturbances are low because it is remote from the continental coast; and 3) water transparency is usually clear because it is remote * Universidad Simon Bolivar, Departamento de Biologia de Organismos, Apartado 89000, Caracas 1080-A, Venezuela. Email: jposada@usb.ve 2 . Fundacion Cientifica Los Roques, Apartado 1 139, Caracas 1010-A, Venezuela. ^ Universidad Central de Venezuela, Instituto de Zoologia Tropical, Apartado 47058, Caracas 1041 -A, Venezuela 532 from any significant runoff. As a consequence, the archipelago has as unusually diverse and luxuriant coral fauna (e.g., Mendez Baamonde, 1978; Croquer and Villamizar, 1998; Garcia, 2001). The archipelago sustains an important finfish and shellfish fishery (i.e., 90% of the Venezuelan spiny lobster production) and tourist activities have increased during the last 10 years. In order to manage the rational use of its commercial and natural resources, the park area has been divided into several special-use zones, e.g., full protection (“no take”), recreation, scientific interest, while commercial and recreational fishing are allowed in the rest of the archipelago. The island of Gran Roque is the most densely populated region of the archipelago (approximately 750 inhabitants) and tourism activities are mostly concentrated there or at the nearest surrounding cays. Fishing activities are restricted to hook-and-line (hand and long lines) all year round and fish traps (spiny lobster season, Nov. 1 to Apr. 31). Skin diving is allowed only for the capture of spiny lobsters, and spearfishing is prohibited throughout the entire archipelago. Fish nets were totally banned in 1994. Studies of the fish in the archipelago have been focused on the evaluation of the biological characteristics of several species, most of which are commercially important (e.g., Hauschild, 1984; Posada et al., 1988; Ortaz et al., 1996). A smaller but still valuable effort has been devoted to creating a complete fish inventory (e.g., Cervigon and Alcala, 1997). The purpose of the present study was to characterize the key elements of the fish community of the archipelago according to the protocol established by the AGRRA program (http://coral.aoml.noaa.gov/agra). The results will provide a baseline of the current status of the fish community which is essential for its proper management. In addition, the results are a valuable basis for regional comparisons and for evaluating subsequent changes in the park after future repeat assessments. METHODS The selection of representative assessment sites in reefs where corals are abundant was based on the two senior authors’ previous knowledge of the area along with advice from Patricia Kramer. Eight of the sites were in shallow water (1-7 m) and five were in depths of 8-12 meters (Table 1). Four sites were located in the fore reef of the southern barrier [Punta de Cayo Sal (PCS), Cayo Sal Sur (CSS), Boca de Cote Somero (BCS) and Boca de Cote Profundo (BCP)], while strong winds and currents forced us to restrict the four surveys of the eastern barrier [Boca de Sebastopol (BS), Barrera Este (BE), Boca del Medio (BM) and Cayo Vapor (CV)] to back-reef habitats (Fig. 1). Also surveyed were four fringing reefs surrounding small cays [Dos Mosquises Sur (DMS), Dos Mosquises Herradura (DMH), Crasqui-La Venada (CRV) and Pelona de Rabusqui (PR)] and one patch reef [Noronqui de Abajo (NOR)]. For a more detailed description of the sites and their benthos, see Villamizar et al. (this volume). The field methodology closely followed the AGRRA Version 2.2 fish protocol (see Appendix One, this volume). All surveys were conducted by two trained observers (Posada and Alvarado). A total of 10 30 m x 2 m belt transects were swum in all reefs but one (PCS), for which there were nine transects (Table 1). Counts of serranids were 533 restricted to species of Epinephelus and Mycteroperca; scarids and haemulids less than 5 cm in length were not tallied. The surveys were conducted between 1000 and 1500 hours except for late afternoon assessments in the two westernmost reefs (DMS and DMH). Consistency training was conducted for a day with natural populations prior to beginning the surveys to ensure agreement on fish sightings, identification, counts, and length estimation. Field identification of fishes was based on Humann (1994). Benthos surveys were conducted at the same time (see Villamizar et al., this volume) but were spaced sufficiently far away to avoid uncontrolled perturbations of fish behavior. Data analysis followed the AGRRA protocol. All Spearman correlation analyses were performed at a significance level of p ^ 0.05. Statistical analyses were made using nonparametric tests due to non-normality of the data, even after transformation. Macroalgal index (% macroalgal relative abundance x macroalgal height in cm) was used as a proxy for macroalgal biomass (see Villamizar et al., this volume). A multidimensional scaling analysis (MDS), based on the Bray-Curtis similarity index (Clarke 1993), was carried out to compare the archipelago’s fish community structure on the basis of the relative abundance of the AGRRA fishes. Additional comparisons were made between our fish abundance data and that provided for Los Roques in the Reef Environmental Education Foundation (REEF)’s expert database on September 29, 1999. RESULTS A total of 1 29 belt transects at 1 3 sites in the archipelago were made during October 4-10, 1999 (Table 1). Fifty-nine of the 70 species in our database of AGRRA fishes were observed during the belt-transect surveys, 1 1 of which were found at all sites: Acanthurus bahianus,A. coeruleus, Haemulon flavolineatum, H. sciurus, Microspathodon chrysurus, Ocyurus chrysurus, Scarus croicensis, S. taeniopterus, S. vetula, Sparisoma aurofrenatum, and S. viride. Two of the western reefs were most numerous in terms of AGRRA species (PCS and DMFl with 36 each), while the fewest species were found in shallow reefs close to Gran Roque (NOR, BM and CV with 20, 22 and 23 species, respectively). Curiously, the reefs with the largest numbers of AGRRA species had relatively low fish densities (67.6 and 91.2 individuals/ 100 m , respectively) and vice versa (Table 1). In general, the community structure of the AGRRA fishes was dominated by herbivorous species (scarids, acanthurids and damselfishes). Sparisoma viride was present in 125 of the belt transects conducted during the current study (97%), followed by Acanthurus coeruleus (88%), Mycrospathodon chrysurus (83%), Scarus vetula (81%) and S. croicensis (81%). Carnivores were well represented by the zooplankton feeder, Ocyurus chrysurus, which was observed in 71% of the transects; the only large-sized serranid present in the surveys was Mycteroperca bonaci. When the data for all sites were pooled, the average density by family was highest for scarids (41.0 individuals/ 100 m ) and acanthurids (22.5 individuals/ 100 m ) and lowest for balistids and pomacanthids (each 0.8 individuals/ 100 m ) and serranids (0.9 individuals/100 m^) (Table 2). Scarids clearly predominated at both depths (Fig. 2). 534 Figure 2, Mean fish density (# individuals/100 ± sd) for AGRRA fishes by depth in (A) shallow and (B) deeper sites in Los Roques, Venezuela. Others = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, and Spyraena barracuda. However, both the relative abundance and the density of the scarids and acanthurids decreased with depth while the opposite trend was seen in the carnivorous serranids and lutjanids (Fig. 3). Overall, the densities of the AGRRA fishes (especially the herbivores) were highest in the reefs (many of which were shallow) located in or near the eastern barrier. Average fish densities were lower at the deeper reefs in the southern barrier where carnivorous lutjanids and serranids were more common. Scarus croicensis was the numerical dominant in seven reefs, with a relative abundance that ranged between 1 1.0% (DMH) and 37.2% (DMS). Acanthurus coeruleus predominated at three shallow reefs (BCS, BE and NOR), with relative abundances of 19.1, 15.6 and 15.3%, respectively. Microspathodon chrysurus Acanthurus bahianus showed the highest relative abundance in two sites (20.9% in BS and 22.4% in CV). In 535 the only reef without a dominant herbivore (CSS), a lutjanid, Ocyurus chrysurus, had the highest relative abundance (18.4%). 70,0. Figure 3. Mean fish density (# individuals/100 ± sd) for AGRRA fishes by geographic location in (A) eastern and (B) southern sites in Los Roques, Venezuela. Others = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus and Spyraena barracuda. In the multidimensional ordination plot (Fig. 4), the fish communities in five (CV, BE, BCS, NOR and BS) of the shallow reefs clustered fairly well together and were separated from most (four/five) of the deeper reefs (CSS, PCS, BCP and DMS). Two of the deeper reefs (CSS and PCS), both located in the southern barrier, also appeared to form a separate subcluster. An inverse, but significant correlation was observed between the total density of all the AGRRA fishes and coral reef complexity as represented by the percentage of live 536 1.8 r 1.4 0.6 0.2 -0.2 -0.6 -1.0 -1.4 O Shallow sites Deep sites BCS BE O O CV o BS o NOR o -0.6 CSS « DMH o BM CRV o -0.2 0.2 Dimension 1 BCR ♦ PR o PCS DMS 0.6 1.4 Figure 4. Multidimensional ordination plot of AGRRA fish transect data in Los Roques, Venezuela. Stony coral cover (r = -0.33; p = 0.30). However, there was no relationship between total AGRRA fish density and the diameter (r = 0.54; p = 0.05) or height (r = 0.29; p = 0.37) of the “large” (>25 cm diameter) stony corals nor between the ratio of these two parameters of reef rugosity (r = 0.36; p = 0.24). Nor was there any relationship between the total density of the key AGRRA herbivores and macroalgal index (r = 0.33; p = 0.30). Significant correlations were found between the density of individual fish families and several coral reef parameters. There was a clear inverse relationship between the density of acanthurids and total live stony coral cover (r = -0.61; p = 0.026) as well as the number of species (r = -0.84; p = 0.01) and density (as number/10 m) (r = -0.69; p = 0.009) of the large stony corals. Lutjanid density was significantly correlated with the number of large eoral species (r = 0.69; p = 0.01) while the density of serranids was significantly correlated with large coral density (r = 0.893; p < 0.0001) and total live stony coral cover (r = 0.80; p = 0.0009). Size frequency distributions of the AGRRA carnivores (Epinephelus, Mycteroperca and all lutjanids) and herbivores (scarids >5 cm, acanthurids, Microspathodon chrysurus) are shown in Figure 5. The size class that was 1 1-20 cm in total length (TL) dominated the length frequency distribution when all species of AGRRA fishes in all sites were grouped. Similar results were observed when subsets of the data were examined by depth stratum or reef type. DISCUSSION The list of species included in our AGRRA database appears to be adequate for a rapid characterization of the key elements of the Los Roques reef fish community (Table 4). For example, nearly 50% of the 59 species observed in the present belt-transect study were ineluded in the 50 most common species listed in the REEF geographic report for Los Roques in September, 1999. 537 O c 3 CT O k. tu Length class (TL, cm) Figure 5. Size frequency distribution of herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) and carnivores (all lutjanids, select serranids) in Los Roques, Venezuela. Cervigon and Alcala (1997) listed an accumulated total of 307 fish species for the Los Roques archipelago, similar to the 306 species reported from 547 Bonaire expert REEF surveys and clearly above the 270 species found in 777 Key Largo expert REEF fish counts (http://www.reef org). The high fish diversity found recently by REEF surveys at Los Roques archipelago (217 species, expert data), despite the relatively low effort invested (106 surveys), is an unequivocal sign of an intact fish community. Herbivores that were widespread in the AGRRA surveys (S. viride, A. coeruleus, M. chrysurus, S. vetula, S. croicensis) have previously been reported as common and highly abundant in both coastal and insular Venezuelan reef systems (Cervigon, 1993, 1994), while Ocyurus chrysurus represents the major finfish target in a fishery that, at Los Roques, is oriented towards spiny lobsters (Posada and Brunetti, 1988). Scarids and acanthurids are lightly harvested since the prohibitions against trapping, spearing, and netting of reef fishes are enforced by local authorities. In consequence, the archipelago may be considered to be a nearly unfished reef fish population. The predominance of herbivorous species in 12 of the 13 Los Roques assessment sites can be associated with good water quality (high transparency) which facilitates benthic algal photosynthetic activities. However, abundance of herbivorous fishes declines, as expected, with depth (Choat, 1991) as the abundance of carnivores increases. That one reef (CSS) was dominated by a carnivore {Ocyurus chrysurus) is likely due to its proximity to the reef edge and the strong currents, full of transient zooplankton, of the southern barrier. A large number of fish species is often related to a complex coral reef topography (Roberts and Ormond 1987). However, contradictory results were observed between fish density and different measurements of reef complexity examined at Los Roques. While significant inverse relationships were observed between total AGRRA fish density and the percentage of live stony coral cover (which is not directly related to topography), there were no relationships with other indicators of reef rugosity (i.e., diameter or height 538 of the large corals, or the ratio of these two parameters). Probably none of these simple parameters are adequate indicators of optimum fish habitat and/or other processes (food availability, water quality, wave action, recruitment, community interactions; Williams, 1991) are simultaneously influencing fish abundance. In our assessments, total fish density varied between 70.8 and 124 individuals/ 100 m^. These values are below those for quantitative fish surveys reported by Alvarado (2000) at Morrocoy National Park (101-287 individuals/ 100 m^), a protected coral reef located in Venezuela’s central coastal zone. However, because Alvarado (2000) modified the AGRRA protocol to include all species of the Pomacentridae and Labridae, the difference between these two estimates is not considered significant. Further comparisons between these two national parks could be enlightening since the two areas differ in the degree to which each is exposed to perturbations. In contrast to Los Roques, Morrocoy National Park is located in a region highly influenced by river runoff and by a greater influence of anthropogenic activities and intense fishing pressure. In addition, its coral reefs were affected in 1996 by an as-yet-undetermined mass mortality event which, in some areas, has decreased the percentage of live stony coral cover from 40% to 1 % and increased the proportion of macroalgae from 1 1% to 55% (Losada and Klein, unpublished manuscript). Overall fish densities for the Scaridae, Acanthuridae, and Serranidae were very similar in Los Roques (Table 2) and Morrocoy (39.0, 22.0 and 1.0 individual/100 m , respectively; Alvarado, 2000). However, lutjanids were significantly more abundant in Los Roques than in Morrocoy (10.7 individuals versus 3 individuals/ 100 m , respectively). The low proportion of carnivorous species in the Morrocoy National Park fish community can be attributed to more intense fishing pressure and its smaller area. Fish community structure was also dominated by herbivorous fishes in the Morrocoy National Park. However, whereas Scarus croicensis dominated in all eight of Alvarado’s (2000) survey sites at Morrocoy, dominance in the Los Roques archipelago was distributed among several species {Scarus croicensis, Acanthurus bahianus and A. coeruleus) and families (Scaridae, Acanthuridae). In Kenya, the predominance of just one species of scarid has been linked by McClanahan (1994) to areas not protected from fishing activities. While the 1 1-20 cm TL size class dominated the size frequency distributions in most of the fish families surveyed at both Los Roques and Morrocoy National Parks (Alvarado, 2000), there was a slight tendency for proportionately higher numbers of individuals to occur in the smaller (6-10 cm TL) size class in Morrocoy. Size classes above 20 cm TL were better represented in the Los Roques data, particularly for commercial species of carnivores. Differences in size structure between these two populations can be attributed to fishing pressure as it is well known that the larger individuals in a population tend to be targeted by most fishing techniques used to capture coral reef species (Russ, 1991). In general terms, the reef fish communities in the Archipielago de Los Roques National Park appear to be in a healthy state (highly diverse community, an abundance of individuals, some large-sized species, a good balance between herbivores and carnivores, etc.) in close concurrence with the good conditions of most of the coral reef sites visited during the present study (see Garcia, 2001; Villamizar et al., this volume). Overall, the key elements of the fish community were more balanced in the deeper southern barrier and southwestern fringing reefs. Fewer commercially important carnivorous fishes were found close to Gran Roque and the eastern barrier. The cause(s) of this disparity could be 539 related to natural differences associated with geographic location or depth, or could possibly be the result of anthropogenic impacts from the islands. It is recommended that special attention be given to these northeastern and eastern areas in any future studies. Thanks to its effective protection by the Institute Nacional de Parques (fNPARQUES), the Los Roques archipelago provides one of the few opportunities in the wider Caribbean to examine a minimally disturbed fish community. This feature is of crucial importance since it provides a baseline standard against which population parameters may be measured in areas that are fished. According to H. Choat and R. Robertson (personal communication), there are few reef systems in the Caribbean with comparable high abundances of large herbivorous and carnivorous fishes. Hence, the Los Roques fish populations should be periodically monitored (at least biannually) in order to become aware of any declines. Research would also be needed to determine the factor(s) that are potentially responsible for these changes and to respond with a contingency plan. Rapid assessments, as provided by the AGRRA protocols, are very useful in countries where economic support for long-term monitoring studies is limited. ACKNOWLEDGMENTS The authors would like to thank the Caribbean Environment Programme of the United Nations Environment Programme for financial support of this study. The AGRRA Organizing Committee facilitated this support, including our participation in the workshops, as described in the Forward to this volume. We are grateful to fNPARQUES for permission to conduct the surveys at Los Roques and to Fundacion Cientifica Los Roques for providing access to their research facilities at Dos Mosquises. Judith Lang and two anonymous reviewers made constructive comments and helpful advice. REFERENCES Alvarado, D. 2000. Variabilidad Espacial de la Estnictura de la Comunidad de Peces de Arrecifes del Parque Nacional Morrocoy. Trabajo Especial de Grado. Universidad Simon Bolivar. 81 pp. Cervigon, F. 1993. Los Peces Marinos de Venezuela (Volumen II). Fundacion Cientifica Los Roques. Caracas. 497 pp. Cervigon, F. 1994. Los Peces Marinos de Venezuela (Volumen III). Editorial Ex Libris, Caracas. 295 pp. Cervigon, F., and A. Alcala 1 997. Peces del Archipielago de Los Roques. Fundacion Museo del Mar y Fundacion Cientifica Los Roques. Caracas, Venezuela. 79 pp. Choat, J.H. 1991. The biology of herbivorous fishes on coral reefs. Pp. 120-1 55. In: P.F. Sale, (ed.). The Ecology of Fishes on Coral Reefs. Academic Press, San Diego, California. 540 Clark, K. 1993. Non-parametric multivariated analyses of changes in community structure. Australian Journal of Ecology 18:117-143. Croquer, A., and E. Villamizar 1 998. Las variaciones de la pendiente topografica, un factor a considerar en la evaluacion de la estructura de una comunidad arrecifal. Revista de Biologla Tropical 46:29-40. Garcia, A. 200 1 . Enfermedades y Otras Anomallas que Afectan a los Corales Escletactldidos en Algunas localidades del Par que Nacional Archipielago de Los Roques. Trabajo Especial de Grado. Escuela de Biologia, Universidad Central de Venezuela. 169 pp. Hauschild, M. 1984. Algunos aspectos de la biologia del carite, Scomberomorus regalis, en el suroeste del Archipielago de Los Roques. Acta Cientlfica Venezolana 35:474. Elumann, P. 1994. Reef fish identification Florida, Caribbean, Bahamas. New World Publications, Inc. Jacksonville, Florida. 396 pp. McClanahan, T.R. 1994. Kenyan coral reef lagoon fish: effects of fishing, substrate complexity and sea urchins. Coral Reefs 13:231-241. Mendez Baamonde, J. 1978. Archipielago Los Roques/Isla de Aves. Cuademos Lagoven. Caracas, Venezuela. 48 pp. Ortaz, M., M.E. Rocha, and J.M. Posada 1996. Food habits of the sympatric fishes, Harengula humeralis y Harengula clupeola (Clupeidae) in the Archipielago de Los Roques National Park, Venezuela. Caribbean Journal of Science 32:26-32. Posada, J.M., and E. Brunetti 1988. Analisis del sistema pesquero del Parque Nacional Archipielago de Los Roques. Caracterizacibn general de la pesqueria. Memorias de la Sociedad de Ciencias Naturales La Salle 3:461-477. Posada, J.M., B. Alvarez, and J.A. Marval 1988. Algunos aspectos de la biologia reproductiva de las sardinas Harengula humeralis y Harengula clupeola en el Parque Nacional Archipielago de Los Roques. Memorias de la Sociedad de Ciencias Naturales La Salle 3:237-258. Roberts, C.M., and R.F.G. Ormond 1987. Habitat complexity and coral reef fish diversity and abundance on Red Sea fringing reefs. Marine Ecology Progress Series. 41:1-8. Russ, G.R. 1991. Coral reef fisheries: effects and yields. Pp. 601-635. In P.F. Sale (ed.). The Ecology of Fishes on Coral Reefs. Academic Press. San Diego, California. Williams, D. McB. 1991. Patterns and processes in the distribution of coral reef fishes. Pp. 437-474. In P.F. Sale (ed.). The Ecology of Fishes on Coral Reefs. Academic Press. San Diego, California. Table 1. Site information for AGRRA fish surveys in the Archipielago de Los Roques National Park, Venezuela. 541 !-i cd CLh .2 c/2 -I *0 C3 CQ C3 -a a c3 E o cu O c3 •o O •o o a sz U OJ C3 if C lO S Al C3 :i: s: ■is ■S * sl' h u ^ c -o E •r CJ All shallow 29.7 ± 18.65 45.0 ± 14.34 7.6 ±7.19 11.3 ± 5.59 7.9 ±4.61 0.4 ± 0.45 2.8 ± 2.28 0.5 ±0.61 0.9 ± 1.59 857.2 Alldeep 11.0±9.50 34.8± 19.33 5.9±5.01 5.0±4.10 15.3±7.42 1.8±1.25 5.4 ± 2.02 1.3±0.66 0.5±0.35 408.5 All sites I 22.5 ± 17.96 41.0 ± 16.47 6.8 ±16.47 8.9 ± 5.82 10.7 ±6.69 0.9 ± 1.05 3.4 ± 2.47 0.8 ± 0.71 0.8 ± 1.25 1265.7 ^ Epinephelus spp. and Mycteroperca spp. 542 543 Table 4. Fifty most frequently sighted fish species in September ,1999 (REEF geographic data) for the Archipielago de Los Roques National Park, Venezuela, with densities (mean ± standard deviation) for species recorded in belt transects (present study). Scientific name Sighting frequency (%)" Density (#/100mh Scientific name Sighting frequency (%) Density (#/100mh) Acanthurus bahianus (.)' 97 7.4 ±8.5 Chaetodon striatus (») 80 0.6 ±0.5 Sparisoma viride (.) 95 9.0 ±3.6 Lactophrys triqueter 79 Ocyurus chrysurus (») 94 5.7 ±4.6 Coryphopterus 78 personatus/hyal in us Acanthurus coeruleus (.) 94 11.1 ± 8.4 Epinephelus cruentatus (.) 77 0.6 ± 1.0 Stegastes partitus 93 Haemulon sciurus (.) 77 1.3 ± 1.1 Halichoeres garnoti 93 Haemulon plumieri (.) 77 0.2 ±0.5 Chromis multilineata 92 Gramma loreto 76 Haemulon flavolineatum (.) 92 3.0 ±2.2 Scarus taeniopterus (.) 75 2.8 ±3.2 Chromis cyanea 92 Caranx ruber (.) 74 0.4 ±0.6 Chaetodon capistratus (*) 92 2.9±2.1 Abudefduf saxatilis 73 Aulostomus maculatus 92 Acanthurus chrirurgus (•) 72 3.9 ±3.7 Thalassoma bifasciatum 91 Holocentrus rufus 72 Holacanthus tricolor (.) 91 0.4 ±0.5 Epinephelus fulvus (.) 71 0.2 ±0.3 Sparisoma aurofrenatum (.) 89 2.3 ± 1.6 Scarus croicensis (•) 70 15.4± 12.9 Clepticus parrae 87 Sphyraena barracuda (.) 66 0.1 ±0.2 Mulloidichthys martinicus 87 Cantherhines macrocerus 66 0.1 ±0.2 Serranus tigrinus 86 Inermia vittata 64 Lutjanus apodus (.) 84 1.9± 1.8 Halichoeres maculipinna 63 Lutjanus mahogoni (.) 82 1.8± 1.8 Haemulon chrysargyreum 62 3.8±3.1 Myripristis jacobus 82 Stegastes planifrons 62 Scarus vetula (.) 82 8.1 ±6.2 Scarus coeruleus (.) 59 0.5± 1.1 Bodianus rufus (.) 82 0.5 ±0.5 Coryphopterus lipernes 58 Canthigaster rostrata 82 Caranx latus 57 Coryphopterus 81 Scarus coelestinus (*) 56 0.4 ±0.6 glaucofraenum Microspathodon chrysurus (») 80 7.0 ±6.5 Hypoplectrus puella 56 'Asterisks indicate the species is listed in the AGRRA protocol. The sighting frequencies for Los Roques as posted at the REEF expert database in September 1999. 544 Frtnchcap /• S o**-' scale in km Caribbean Sea Figure 1. AGRRA survey sites in the Virgin Islands. (1) Caret Bay, (2) Brewer’s Bay, (3) Flat Cay, (4) Buck Island, (5) Sprat Bay, (6) Cane Bay, (7) Salt River, (8) Long Reef, (9) Fish Bay east outer, (10) Fish Bay west outer, (11) Fish Bay east inner, (12) Great Lameshur Donkey, (13) Great Lameshur VIERS, (14) Fish Bay west inner, (15) Great Lameshur Tektite, (16) Great Lameshur Yawzi, (17) Iguana head, (18) Eustatia Reef, (19) Herman’s Reef, (20) Horseshoe Reef, (21) Jack Bay, (22) West Cow Wreck. A RAPID ASSESSMENT OF CORAL REEFS IN THE VIRGIN ISLANDS (PART 1: STONY CORALS AND ALGAE) BY RICHARD S. NEMETH/ ADAM QUANDT,' LAURIE REQUA,' J. PAIGE ROTHENBERGER,^’^ and MARCIA G. TAYLOR^ ABSTRACT The Atlantic and Gulf Rapid Reef Assessment (AGRRA) benthos protocol was conducted in depths of 3-14 m at 22 coral reef sites in the U.S. Virgin Islands (St. Croix, St. Thomas, St. John) and the British Virgin Islands (Anegada, Guana Island and Virgin Gorda). Live stony coral cover averaged between 10% and 35% in 85% percent of the sites. The size of colonies >25 cm in diameter averaged 55 cm and their composition was dominated by the genus Montastraea. Coral recruitment varied considerably among sites and was dominated by species that brood their larvae. Nearly all sites had stony corals that were affected by disease, bleaching or damaged by fish bites. Mean values for total (recent + old) partial mortality exceeded 40% of colony surfaces in eight sites and were between 20% and 40% for the remainder. The abundance of “standing dead” stony corals was typically less than 1.5%. The relative abundance of macroalgae exceeded 30% in 15 of 22 sites and macroalgae were dominated by Dictyota. INTRODUCTION The Virgin Islands (18°20’ N, 64°50’W) lie between two major island archipelagoes: the Greater Antilles to the west and the Lesser Antilles to the southeast. With the exception of St. Croix, the northern United States Virgin Islands (USVI), together with the British Virgin Islands (BVI) and the islands of Puerto Rico, rise from a geological shelf that is surrounded by deep water (Dammann, 1969). This shelf covers ■j approximately 3,200 km , contains about 500 km of shelf edge and is generally less than 100 m in depth. St. Croix sits on a similar, but smaller, shelf that is separated from the ' Center for Marine and Environmental Studies, University of the Virgin Islands, MacLean Marine Science Center, 2 John Brewer's Bay, St. Thomas, UVSI 00802-9990. Email: memeth@uvi.edu 2 Center for Marine and Environmental Studies, University of the Virgin Islands, Virgin Islands Marine Advisory Service, PO Box 10000, Kings Hill, St. Croix USVI 00850. ^ Department of Environmental Science and Policy, George Mason University, 4400 University Dr., MSN 3E1, Fairfax, VA 22030. 546 northerly Virgin Islands by the 4,685 m deep, 60 km wide Virgin Islands trough. Upwelling of deep, nutrient-rich waters from the North Equatorial and Caribbean Currents, coupled with the fact that these shelves lie within the photic zone, aid in the development of deep (45 m) and shallow (2 to 20 m) fringing reefs, bank barrier reefs and patch reefs (Dammann, 1969). St. Thomas, St. John and St. Croix, the three main islands of the US VI, are surrounded by more than 90 uninhabited islands, cays, and rocks which provide hard substrata suitable for the growth of coral reefs (Towle, 1970, unpublished report; Dammann and Nellis, 1992). The three main islands of the BVI, Tortola, Virgin Gorda and Anegada, are also surrounded by nearly 50 smaller islands, cays and rocks. With the exception of Anegada and St. Croix, the Virgin Islands are of volcanic origin. Anegada is a low coral island (Dunne and Brown, 1976, unpublished report) and St. Croix has a sedimentary and carbonate origin (Hubbard, 1989; R. Watlington, personal communication). Many of the reefs in the Virgin Islands are associated with mangrove forests, seagrass beds and algal plains, which are important to reef-associated fishes. Over the past 20 years, eight major hurricanes, numerous outbreaks of disease, and sporadic bleaching events have caused extensive coral mortality to the coral reefs surrounding the Virgin Islands (Gladfelter, 1982; Edmunds and Whitman, 1991; Rogers et ah, 1991; Causey et ah, 2000). Recovery from these natural disturbances is hindered by a multitude of human impacts that affect coral reefs such as overfishing, ship groundings, anchor damage and non-point source pollution (Roberts, 1993; Sebens, 1994; Rogers and Garrison, 2001). Moreover, rapid development of inland and coastal areas has dramatically increased soil erosion and sedimentation onto many of these coral reefs (Rogers, 1990; MacDonald et ah, 1997; Anderson and MacDonald, 1998; Ramos, 1998, unpublished report). The cumulative effects of these human impacts reduce coral abundance, diversity, and larval recruitment and may make corals more susceptible to disease and bleaching (Nemeth and Sladek Nowlis, 2001). In response to these threats to coral reefs, Ginsburg et ah (1996) initiated a process of rapid reef assessment. After development of the AGRRA protocols, a Caribbean-wide effort to assess the condition of coral reefs throughout the region was launched. The University of the Virgin Islands’ Center for Marine and Environmental Studies joined the effort and set out to assess the reefs of the Virgin Islands using the AGRRA protocols. This paper reports on the initial findings of our benthic assessments. Results of the fish surveys are given in Nemeth et ah (this volume). METHODS Site-selection criteria in the USVI and the BVI varied among the different islands but most ( 1 8/22) choices were made for strategic reasons. In St. Thomas, five sites were selected based on their inclusion in a long-term, sedimentation monitoring project, their proximity to the University of the Virgin Islands McLean Marine Science Center and the presence of historical data (i.e., Rogers, 1982 , unpublished report; Nemeth and Sladek Nowlis, 2001). Three popular recreational diving sites were selected in St. Croix, one of which is within the Salt River Bay National Historic Park and Ecological Preserve. The eight sites in St. John were part of a study comparing sedimentation rates between Great 547 Lameshur Bay, which is within the National Park Boundary, and Fish Bay, which has been experiencing heavy development within its watershed. Four of these sites were shallow reefs (<6 m) located inside the bays and four sites were in deeper (>6 m) reefs located outside the bays. Four of the six sites surveyed in the BVI (Iguana Head in Guana Island, Eustatia Reef in Virgin Gorda, West Cow Wreck Bay and Herman’s Reef in Anegada) were seleeted haphazardly. Eustatia Reef is a heavily visited dive site whereas Guana Island is a privately owned island with a low human population. The other two sites in Anegada (Jaek Bay and Horseshoe Reef, a designated protected area) were seleeted because of historical surveys eondueted by Dunne and Brown (1976, unpublished report) and the West Indies Laboratory (1983, unpublished report). The reefs of Anegada were included in the AGRRA survey to provide a remote reference site with low human population and little landmass. A qualitative assessment of human and natural impacts at the 22 sites surveyed in the Virgin Islands is given in Appendix A (this paper). The AGRRA Version 2.0 benthos protocol (see Appendix One, this volume) was used. Four of us constituted the primary dive team and we were augmented by five alternates. All divers partieipated in at least one three-day training session of the AGRRA protocol. Training sessions were conducted in the spring of 1998, 1999, and 2000 and consistency training was used for the primary dive team at least once a year. All corals were identified to species except species of Agaricia. Although A. agaricites was a dominant species, it was not distinguished from other species of Agaricia. Coral and algal identification guides included Humann (1993) and Littler et al. (1989). Stony coral sizes were measured to the closest cm. We also recorded the percentage of individually surveyed (>25 cm diameter) colonies with parrotfish or damselfish bites. Fish-bite damage from parrotfish or damselfish was distinguished when possible. Maeroalgal heights were measured to the closest 0.5 cm. When scoring the relative abundance of crustose coralline algae, divers typically removed sediment with vigorous handsweeping motions. Algal turfs were omitted from the July 2000 assessment of Anegada and Virgin Gorda sites in accordanee with the May 2000 revisions of the AGRRA benthos protoeol. Beeause of these ehanges, percent relative abundanees of macroalgae and crustose eoralline algae are not comparable to earlier surveys. Therefore, pereent absolute abundance of macroalgae in the quadrats is presented to make comparisons among the sites. Data were summarized by island groups within three geographic areas to examine general trends in coral reef condition throughout the Virgin Islands. The geographie areas were: 1) Anegada; 2) the shallow and deeper reefs of the remaining islands in the northern Virgin Islands (NVI); and 3) St. Croix. Anegada was eonsidered a geographic unit because of its isolation from the other Virgin Islands and its unique geology (low coral island). St. Croix was considered a geographic unit also because of its isolation from the NVI, its unique geology (sedimentary/carbonate) and because it is eompletely within the Caribbean Sea. The NVI (USVI = St. Thomas, St. John; BVI == Guana, Virgin Gorda) were grouped beeause of their elose proximity to one another, their similar geologie origins and topographies (high volcanic islands) and their exposure to both Atlantic waters from the north and Caribbean waters from the south. The shallow sites in St. John were analyzed separately from the deeper sites within the NVI archipelago. Comparisons of coral cover among “island/depth groups” (hereafter referred to as island groups) were made using single factor Analysis of Variances (ANOVA). Prior to statistical analysis, residuals of eoral cover data were graphically analyzed for normality 548 using Systat statistical analysis software. Coral cover data were found suitable for statistical analysis at a significance level of alpha = 0.05. Tukey HSD was used for post- hoc multiple comparison analyses. RESULTS Site Characteristics A total of 22 sites in eight islands or cays (Fig. 1) were surveyed during three time periods: May 1998; May to December 1999; and February to October 2000 (Table 1). All of the surveyed sites were in fringing reefs at depths that ranged from 3 to 14 m (Table 1). The fringing reefs in St. Thomas, St. John, Guana Island and Virgin Gorda were all located close to the shoreline and sloped relatively steeply from their crests to their bases where the reefs gave way to sandy substrata. In St. Croix, the reefs were within a few hundred meters of the Virgin Islands trough and dropped off steeply into abyssal depths. In the low island of Anegada, the fringing reefs sloped more gently than in the high volcanic NVI. All of the deeper reef surveys were carried out on their seaward slopes with the exception of Herman’s Reef in Anegada which was on the leeward side of the reef crest. Most of the sites that were surveyed were relatively sheltered from the prevailing seas (i.e., from exposure to northeast swells). Sites with direct exposure to northeast swells included Jack Bay, Horseshoe Reef, and West Cow Wreck Bay in Anegada. The three reefs in St. Croix were also exposed to swells although the Virgin Islands archipelago 60 km to the north provided some protection. Reefs with moderate protection from prevailing seas included Herman’s Reef in Anegada, Caret Bay in St. Thomas, and Eustatia Reef in Virgin Gorda. Over 85% of the sites surveyed had between 10% and 35% live stony coral cover (Table 1) with means for the island groups ranging between 12% and 19% (Fig. 2). The lowest cover occurred in Fish Bay, St. John whereas the deeper Tektite site in Great Lameshur Bay, St. John had exceptionally high (nearly 50%) live coral cover (Table 1). Significant differences in percent coral cover occurred among island groups (ANOVA: F3,34i = 12.4, P < 0.001). Anegada and the shallow St. John sites were significantly lower in coral cover than sites in the deeper NVI and St. Croix groups (Fig. 2). The “large” stony corals that were individually surveyed (i.e., those >25 cm in diameter) were numerically dominated by the Montastraea annularis species complex (M annularis, M. faveolata, M. franksi) in all island groups except St. Croix which was dominated by M. cavernosa (Fig. 3). The deeper NVI and shallow St. John sites were similar in overall composition and quite different from those in St. Croix and Anegada, which also differed from each other (Fig. 3). The second most common taxon was Siderastrea siderea, except in Anegada where Porites astreoides and Diploria strigosa were each twice as common as S. siderea. Other individually surveyed taxa that were each less than three percent in abundance in the four island groups are as follows: shallow St. John {Millepora alcicornis, Agaricia, Colpophyllia natans, Diploria labyrinthiformis, D. strigosa, Stephanocoenia intersepta, Solenastrea bournoni)\ deeper NVI (M alcicornis, Acropora cervicornis, A. palmata, Dichocoenia stokesi, Madracis mirabilis, M. decactis, S. intersepta, Dendrogyra cylindrus, Solenastrea bournoni, D. 549 Figure 2. Mean live stony coral cover (percent ± se) in St. Croix, deeper NVI (St. Thomas, St. John >6 m deep. Guana Is., Virgin Gorda), shallow St. John and Anegada. Lines connecting bars indicate no significant difference (a = 0.05). labyrinthiformis, D. strigosa); St. Croix (Z). cylindrus, D. labyrinthiformis, M. decactis, Porites porites); and Anegada {P. porites, M. franksi, A. cervicornis, D. stokesi, S. intersepta, D. clivosa, D. cylindrus, M. mirabilis). The diameter of the individually surveyed corals (Table 2) averaged 55.0 cm with exceptionally large (up to 260 cm) colonies of M. annularis occurring off St. John. Within the Montastraea annularis species complex, the size frequency distribution in most sites showed a positive skew toward smaller colonies of <50 cm diameter (Fig. 4). The shallow St. John sites were unique in that they showed a platykurtic size-frequency distribution (Fig. 4) even though the local abundance of the M. annularis species complex as a whole was similar to that of the deeper NVI sites (Fig. 3). Three of the four shallow sites off St. John had very large coral colonies and so exerted a proportionately greater influence in this group than did the two deeper Great Lameshur sites in the much larger set of the deeper NVI group. The mean sizes of the predominant Montastraea varied within and among the island groups (Fig. 5). M. annularis and M. faveolata were significantly larger than M. franksi and M. cavernosa in the deeper NVI (F3,678 = 13.72, p<0.001) whereas, in St. Croix, M. faveolata and M. franksi were significantly smaller than M. annularis and larger than M. cavernosa (F3,20i = 22.47, p<0.001). In the shallow reefs of St. John, M. cavernosa was significantly smaller than M. annularis (F3,25i = 3.64, p<0.01) whereas M. faveolata and M. franksi were intermediate in size. All Montastraea on Anegada were similar in size (Fig. 5). Stony Coral Condition Coral bleaching was recorded in all sites (Table 2) with the highest average values for large stony corals occurring in St. Croix (48% in October/December 1999, n=3 550 Figure 3. Species composition and mean relative abundance of all stony corals (>25 cm diameter) in (A) shallow St. John, (B) deeper NVI, (C) St. Croix, and (D) Anegada. ap = Acropora palmata, ag = Agaricia, cn = Colpophyllia natans, dl = Diploria labyrinthiformis, ds = D. strigosa, mm = Madracis mirabilis, mil = Millepora alcicomis, ma = Montastraea annularis, mfa = l faveolata, mff = M franksi, me = M cavernosa, pa = Porites astreoides, pp = P. porites, ss = Siderastrea siderea, other = all species with <3% abundance. 551 30 -| 20 - 10 - 0 - 30 n 20 - 10 - o c 0) 3 0 -1 O' CD 30 n Ll_ 20 - 10 - 0 - 30 -| 20 - 10 - ! m' faveo^ta ^ Shallow St. John n=223 M. franksi B Deeper NVI n=558 C St Croix n=82 D Anegada n=112 ^ Coral colony diameter (cm) Figure 4. Size-frequency distribution of all colonies (>25 cm diameter) of Montastraea annularis (ma), M faveolata (mfa) and M. franksi (mfr) in (A) shallow St. John (ma=162, mfa=34, mfr -21), (B) deeper NVI (ma=237, mfa=178, mfr=143), (C) St. Croix (ma=24, mfa=24, mfr=34), and (D) Anegada. (ma=64, mfa= 42, mfr=6). sites), in two shallow St. John sites (~38% in October/November 1999), in St. John’s deeper Tektite Reef (36%, in August/November 1999) and in Anegada (28% in July 2000. n=4 sites). Occasional temperature measurements during surveys documented that normal sea-surface temperatures to 10 m depth were coolest in February (25° C), began to warm during June (26° C) and July (27° C), peaked in August (28-29° C), and began to cool during October and November (27° C). Water temperatures deeper than 12 m could be up to 1.5° C cooler than in shallower depths. The Caribbean-wide 1998 bleaching event was first detected in September 1998 in the Virgin Islands when surface sea-water temperatures exceeded 30° C. Bleaching of coral colonies peaked in October 1998 but nearly all signs of bleaching had disappeared by February 1999 (Nemeth and Sladek Nowlis, 2001). Since no surveys were conducted from June 1998 to May 1999, the 1998 bleaching event did not directly influence the results presented in this report. 552 if) Figure 5. Mean size of Montastraea annularis (ma), M. faveolata (mfa), M franksi (mfr) and M cavernosa (me) in A) shallow St. John (nia=162, mfa=34, mfr =27, mc=32), (B) deeper NVI (ma=237, mfa=178, mfr=143, mc=124), (C) St. Croix (ma=24, mfa=24, mfr=34. mc=123), and (D) Anegada (ma=64, mfa= 42, mfr=6, mc=66). Lines connecting bars indicate no significant differences (a=0.05) in average size among species. Signs of disease in the individually surveyed stony corals were present in most (20/22) of our assessment sites (Table 2). Divers were able to recognize four general disease types: black band, yellow blotch, white plague, and dark spots (Fig. 6). The species most susceptible to disease included Montastrea faveolata, M. franksi, M. cavernosa, M. annularis, Colpophyllia natans and Siderastrea siderea. Of particular interest, white-band disease was observed in one colony of Acropora cervicornis each in Caret Bay, St. Thomas and in Herman Reef, Anegada. The percent of diseased corals 553 Island Group Figure 6. Percent damage to all stony corals (>25 cm diameter) from disease in shallow St. John (n=18 infected corals), deeper NVI (n=106 infected corals), St. Croix (n=6 infected corals), and Anegada (n=8 infected corals). varied among sites within island groups with St. Croix (in October and December 1999) and Anegada (in July 2000) having lower levels of disease than the shallow St. John sites (in October/November 1999 and May/August/October 2000) and deeper NVI sites (various dates between May 1998 and July 2000). Five percent or more of the colonies were affected at 13 sites (maximum being 24% in Guana Island), 9 of which included surveys that had been made between May and November, 1999. At 4 sites in St. Thomas, sampling dates either occurred both in 1998 and 1999 (Buck Island), in 1998 and 2000 (Caret Bay), or in 1999 and 2000 (Brewers Bay, Flat Cay) (Table 1). In the first two cases no diseases were observed in 1998 but they were present in all sites on the other survey dates. The incidence of disease was considerably higher in 1999 than in 2000 for Brewer’s Bay. Damselfish and parrotfish bites were responsible for causing tissue damage to 25% or more of the individually surveyed stony corals in three sites and were present in colonies at all but three sites (Table 2; Fig. 7). The most frequently attacked species included M. annularis, M. faveolata, M. franksi, C. natans, M. cavernosa and P. porites. The shallow sites in Great Lameshur Bay, St. John had the greatest percentage of tissue damage from damselfish. This was largely due to the high density of three-spot damselfish (Stegastes planifrons) on colonies of M annularis. Damselfish overall occupied 35.4% of the stony corals in the shallow St. John sites, 7.1% in the deeper NVI, 3.2% in Anegada and 2.0% in St. Croix. Tissue damage by parrotfish was greatest in St. Croix and Anegada (Fig. 7). Mean values for total (recent and old) partial mortality exceeded 40% of colony surfaces in eight sites and were between 20% and 40% in the remaining sites (Table 2). Recent partial mortality of colony surfaces averaged less than three percent overall in 554 20 15 c (D o 10 (U Oh 5 0 shallow SJ BwgB unid.fish bite parrotfish bites damselfish bites Island Group Figure 7. Percent damage to all stony corals (>25 cm diameter) from fish bites in shallow St. John (71 corals with bites), deeper NVI (60 corals with bites), St. Croix (20 corals with bites), and Anegada (48 corals with bites). most (17/22) sites (Table 2). Although the highest rate of recent partial mortality (5.5%) for an individual reef was in Anegada (Jack Bay, Table 2), the highest rates of recent partial mortality for an island group occurred in the shallow St. John sites where over 50% of the large stony corals were affected during the surveys in 1999 and 2000 (Fig. 8). Fifty-five to 70% percent of the corals throughout the Virgin Islands had signs of old tissue mortality on more than 10% of their upper surface areas (Fig. 8). Average levels of old tissue mortality per coral colony were highest in the shallow St. John sites (43%) followed by Anegada (40%), St. Croix (34%) and the deeper NVI (27%). The percent of large colonies that were “standing dead” (no living tissues and still in growth position) was less than or equal to 1.5% at 17 sites but was relatively high (8% and 12%) in two reefs off Anegada (Table 2). The average frequency of standing dead colonies by island group was as follows: St. Croix (0.3%, n=l coral), deeper NVI (0.3%, n=4 corals), shallow St. John (1.7%, n=7 corals), and Anegada (9%, n=36 corals). Standing dead corals in shallow St. John, deeper NVI and St. Croix consisted of a variety of species: M. cavernosa, M. faveolata, M. annularis, S. siderea, P. astreoides, C. natans, and A. palmata. In Anegada, however, all (12/12) of the standing dead colonies at Horseshoe Reef were A. palmata and most (1 1/14) of the standing dead colonies at Jack Bay were M cavernosa. Other standing dead colonies at Anegada included Agaricia, D. labyrinthiformis, D. strigosa, M. annularis, M. franksi, P. astreoides, and P. porites. 555 100 n 80 - 60 - 40 - 20 - 1 ZZD 1 % old mortality % new mortality A Shallow St. John n=407 u 100 n 80 - 60 - ^ 40- S 20 - ^ n 1 1 B Deeper NVI n=l,190 l_ 0 L— 1 100 n O 80- ^ 60 - 40 - 20 - ■ 1 C St. Croix n=301 niaaa.B.BB u ^ 100 - 80 - 60 - 40 - 20 - ■ 1 D Anegada n=404 ■ „B-aB>iHBflB U 1 1 1 1 1 1 1 1 1 1 1 ^ ^ ^ ^ c? ^ 25 cm diameter) in (A) shallow St. John, (B) deeper NVI, (C) St. Croix, and (D) Anegada. Stony Coral Recruitment, Macroalgae and Diadema Stony coral recruitment varied considerably from site to site (Table 3). With the exception of Siderastrea siderea, coral recruits were dominated by species that brood their larvae. The five most abundant taxa, Siderastrea siderea (23%), Agaricia (17%), P. astreoides (15%), P. porites (13%J and S. radians (6%) comprised 70% to 80% of the recruits in all island groups (Fig. 9). Other recruits that were each less than three percent in abundance in the four island groups are as follows: shallow St. John {D. strigosa, D. labyrinthiformis, M. mirabilis); deeper NVI (M mirabilis, S. bournoni, M. cavernosa, D. 556 A Shallow St. John n=59 Other 3% ff3% B Deeper NVIn=3 08 Other 3% ff 3% ; C St Croix n=116 D Anegada n=89 Figure 9. Species composition and mean relative abundance of all stony coral recruits (<2 cm diameter) in (A) St. John shallow, (B) Northern Virgin Islands, (C) St. Croix, (D) Anegada. AG = Agaricia, ds = Diploria strigosa, ff= Favia fragum, md = Madracis decactis, mil = Millepora alcicornis, ma = Montastraea annularis species complex, me = Montastraea cavernosa, pa = Porites astreoides, pp = P. porites, sr = Siderastrea radians, ss = S. siderea, sm = Stephanocoenia intersepta, sb = Solenastrea bournoni. Other = all species with <3% abundance. 557 labyrinthiformis, A. cervicornis, D. stokesi, Eusmillia fastigiata, S. intersepta, D. clivosa, M. alcicornis)\ St. Croix (M mirabilis, S. intersepta, D. stokesi, E. fastigiata, M. alcicornis, F. fragum, M. annularis)’, and Anegada (Z). labyrinthiformis, M. alcicornis, M. annularis, M. cavernosa). Of the 16 sites in which relative abundance estimates were made for all three algal functional groups, four were dominated by macroalgae, eight were dominated by turf algae and these two groups were essentially coequals in the remaining four. Crustose coralline algae were the least abundant at all but one site (Caret Bay, St. Thomas) where they were more abundant than turf algae. When comparing absolute abundances of macroalgae within the quadrats, over half (14/22) of the sites ranged from 25% to 45%, two were over 50% and three sites were under 10% (Table 3). Dictyota was the dominant macroalga in the Virgin Islands reefs and macroalgae were typically less than 3 cm in height (Table 3). Values for the macroalgal index (as relative macroalgal abundance x macroalgal height) obtained from the 1998-1999 surveys ranged from 3-35 in St. Croix, 6-247 in the deeper NVI and 46-278 in the shallow St. John sites. Corresponding values based on absolute macroalgal abundances for the entire dataset produced a similar pattern with shallow St. John>Anegada>deeper NVI>St. Croix (Table 3). The density of Diadema antillarum was greatest in the shallow St. John sites (i.e., in depths of less than five meters) (Table 3). DISCUSSION It was anticipated that coral reefs located near human population centers would be exposed to higher levels of sedimentation, pollution, recreational diving, anchor damage and more intense fishing and thus would be in worse condition than reefs more remote from these human impacts (Appendix A, this paper). Contrary to this assumption, our surveys found that the Anegada sites had the greatest amount of dead stony coral tissue and the lowest live coral cover among the deeper reef systems. Old mortality in Anegada was largely attributed to the fact that its reefs were historically dominated by Acropora palmata which was decimated during a white-band disease epizooitic in the late 1970’s (Dunne and Brown, 1976, unpublished report; Gladfelter, 1982). During our survey, many of these A. palmata were broken or severely eroded with other species of corals colonizing their surfaces; thus many did not get categorized as standing dead colonies. The shallow St. John sites were fairly similar to the deeper NVI sites in the species composition of large corals and in coral recruitment (Figs. 3 and 9). However, the shallow St. John sites had a much higher percentage of old partial-colony mortality than the deeper NVI (43% vs. 27%) and had fewer small (<50 cm diameter) colonies of the Montastraea annularis species complex (29% vs. 50%) (Fig 4). The partial death or damage to these shallow-water corals was possibly due to a series of strong hurricanes impacting St. John’s reefs during the past 20 years or to possible epizootic events (see below). All else being equal, large storm waves have a greater impact in shallow reefs by breaking branching corals, dislodging or toppling boulder corals, or scouring thin coral tissues (Rogers et al., 1983; Edmonds and Witman, 1991; Rogers et al., 1991; Nemeth and Sladek Nowlis, 2001). The shallow St. John sites also contained higher densities of 558 Diadema antillarum, a pattern of recolonization that is typical throughout the Caribbean (Nemeth, personal observation) following its die-off in the early 1980’s (Lessios, 1988). The shallow reefs also contained higher densities of damselfish, in particular Stegastes planifrons, which is known to inhabit shallow reef areas and to damage and kill living coral tissue for its algal gardens (Itzkowitz, 1977; Kaufman, 1977; Williams, 1979). This resulted in large numbers of coral colonies showing signs of tissue damage from damselfish bites. Bites by parrotfish were also assessed. Since the stoplight parrotfish, Sparisoma viride, is known to cause damage to coral colonies (Bruckner and Bruckner, 1998; Bruckner et al., 2000), we examined the AGRRA fish data collected during our surveys (Nemeth et al., this volume) and found that bite damage seemed to be positively associated to the average size of >5 cm stoplight parrotfish present in reef sites. For example, the St. Croix and Anegada reefs showed the highest incidence of parrotfish bite damage and had the largest stoplight parrotfish observed during the AGRRA fish surveys (29.2 cm and 17.7 cm, respectively). In contrast, the average size of stoplight parrotfish in the deeper NVI and in shallow St. John was 13.1 cm and 9.27 cm, respectively. Although the 1 998 bleaching event did not directly influence the results presented in this report, Nemeth and Sladek Nowlis (2001) found that sedimentation from land development made corals more susceptible to bleaching, especially during the 1998 event. Nemeth and Sladek Nowlis (2001) also documented low-to-moderate levels of seasonal bleaching during July, August and early September, when sea surface temperatures typically reach an annual high, and during the rainy season (October, November, early December) when the influx of terrigeneous sediment into the marine environmental is greatest. This may explain the moderate levels of bleaching observed for all sites in St. Croix (October, December) and Anegada (July), four of five sites in St. Thomas (July/August) and the deeper St. John sites (July/August), three of four shallow sites in St. John (August/ October/November) and the one site in Virgin Gorda (July). The lowest levels of bleaching typically occurred in sites surveyed from February to June. Among the island groups, the incidence of disease was lowest in Anegada in July 2000 and in St. Croix during October and December 1999. Although little is known of the mechanisms of transmission of coral diseases, both these islands are remote relative to those in the NVI. Guana Island was unusual in that it was among the highest in live stony coral cover, among the lowest in macroalgal abundance and in the percentage of total partial-colony mortality, but it also had a considerably higher percentage of diseased corals (24% in 1999) than found at any date in any other site (Tables 1-3). Our AGRRA survey may have documented the initial stages of a disease outbreak at this site in which 65% of the affected corals were in the genus Montastraea. Unfortunately, most of the signs of disease in the Guana Island corals were unfamiliar to us and classified as “unknown.” Alternatively, since Guana Island is privately owed, sparsely populated, and its reefs receive minimal human disturbance, its stony corals may be more capable of resisting or recovering from diseases which may have been present for some time. Outbreaks of white plague type II, which occurred repeatedly in St. John (December 1997, April, May, June; December 1998; August 2000), resulted in considerable mortality to several isolated reefs around St. John including Tektite Reef in Great Lameshur Bay (Miller et al., in press). None of our survey dates in St. John corresponded 559 with these peak periods in disease outbreaks. However, the percent of diseased colonies in our surveys in Tektite Reef (10% in August/November 1999) was similar to that reported by Miller et al. (in press) who found 5-12% diseased by white plague type II during the same time period. Temporal variability in outbreaks of disease may have contributed to the low percentage of disease detected at sites in Anegada and Virgin Gorda which were all sampled during a three-day period in July 2000. Based on our comparison of populated islands with remote islands, we conclude that the condition of coral reefs around the Virgin Islands is primarily determined by large-scale natural disturbances such as hurricanes and disease epidemics. Although many reefs in the Caribbean and other parts of the world were severely damaged by the 1998 mass bleaching event, the corals around the Virgin Islands had largely recovered by 1999. The remoteness of Anegada may benefit the ecology of its reefs under certain circumstances but, following disturbances, it is possible that larval recruitment of major, reef-building corals is limited which may slow the recovery of reefs surrounding this island. ACKNOWLEDGMENTS The AGRRA Organizing Committee facilitated financial support from the United States Department of Interior as described in the Forward to this volume. We thank Guana Island Resort and Lianna Jarecki for supporting the UVI assessment team while at Guana Island. This large-scale assessment could not have been accomplished without our dedicated alternate divers: Piero Mannoni, Brianne Loran, Longin Kaczmarsky, Nicole Kellum, Joel Cameron and Stuart Ketchum. Thanks also for the support of our dive officer and captain, Steve Prosterman. REFERENCES Anderson, D.M., and L.H. MacDonald 1998. Modeling road surface sediment production using a vector geographic information system. Earth Surface Processes and Landforms 23:95-107. Bruckner, A.W., and R.J. Bruckner 2000. Destruction of coral by Sparisoma viride. Coral Reefs 17:350. Bruckner, A.W., R.J. Bruckner, and P. Sollins 2000. Parrotfish predation on live coral: “spot biting” and “focused biting.” Coral Reefs 19:50. Causey, B., J. Delaney, E. Diaz, D. Dodge, J.R. Garcia, J. Higgins, W. Jaap, C.A. Matos, G.P. Schmahl, C. Rogers, M.W. Miller, and D.D. Turgeon 2000. Status of coral reefs in the US Caribbean and Gulf of Mexico: Florida, Texas, Puerto Rico, U.S. Virgin Islands and Navassa. Pp. 239-259. In: C. Wilkinson, (ed.). Status of Coral Reefs of the World: 2000. Australian Institute of Marine Science, Cape Ferguson, Queensland and Dampier, Western Australia. 560 Dammann, A.E. 1969. Study of the fisheries potential of the Virgin Islands. Special Report. Virgin Islands Ecological Research Station, University of the Virgin Islands, Caribbean Research Institute. Contribution No. 7. 197 pp. Dammann A.E., and D.W. Nellis 1992. A Natural History Atlas to the Cays of the US Virgin Islands. Pineapple Press, Inc. Sarasota, FI. 160 pp. Edmunds, P.J., and J.D. Witman 1991 . Effect of Hurricane Hugo on the primary framework of a reef along the south shore of St. John, US Virgin Islands. Marine Ecology Progress Series 78:201-204. Ginsburg, R.N., R.P.M Bak., W.E. Kiene, E. Gischler, and V. Kosmynin 1996. Rapid assessment of reef condition using coral vitality. Reef Encounter 19:12-14. Gladfelter, W.G. 1982. White-band disease in Acropora palmata: implications for the structure and growth of shallow reefs. Bulletin of Marine Science 32:639-643. Hubbard, D.K. 1989. Modem carbonate environments of St. Croix and the Caribbean: a general overview. Pp. 85-94. In D.K. Hubbard (ed.). Terrestrial and Marine Geology of St. Croix, U.S. Virgin Islands, Special publication of the West Indies Laboratory, St. Croix 8. Humann, P. 1993. Coral reef identification: Florida, Caribbean, Bahamas. N. DeLoach (ed.). New World Publications, Inc. Jacksonville, Florida. 239 pp. Itzkowitz, M. 1977. Spatial organization of the Jamaican damselfish community. Journal of Experimental Marine Biology and Ecology 68:327-359. Kaufman, L.S. 1977. The threespot damselfish: effects on benthic biota of Caribbean coral reefs. Proceedings of the Third International Coral Reef Symposium 1:559-564. Lessios, H.A. 1988. Mass mortality of Diadema antillarium in the Caribbean: What have we learned? Annual Review of Ecology and Systematics 19:371-393. Littler, D.S., M.M. Littler, K.E Bucher, and J.N. Norris 1989. Marine plants of the Caribbean, a field guide from Florida to Brazil. Smithsonian Institution Press: Washington, D.C. 272 pp. MacDonald, L.H., D.M. Anderson, and W.E Dietrich 1997. Paradise threatened: land use and erosion on St. John, US Virgin Islands. Environmental Management 2 1 :85 1 -863. Miller, J., C. Rogers, and R. Waara In press. Monitoring the coral disease, plague type II, on coral reefs in St. John, U.S. Virgin Islands. Revista de Biologia Tropical. Nemeth, R.S., and J. Sladek Nowlis. 200 1 . Monitoring the effects of land development on the near-shore reef environment of St. Thomas U.S. Virgin Islands. Bulletin of Marine Science 69:759-775. 561 Roberts, C.M. 1993. Coral reefs: health, hazards and history. Trends in Ecology and Evolution 8:425-427. Rogers, C.S. 1990. Responses of coral reefs and reef organisms to sedimentation. Marine Ecology Progress Series 62:185-202. Rogers, C.S., and V.H. Garrison 2001 . Ten years after the crime: Lasting effects of damage from a cruise ship anchor on a coral reef in St. John, U.S. Virgin Islands. Bulletin of Marine Science 69:793-804. Rogers, C.S. M. Gilnack, and H.C. Fitz 1983. Monitoring of coral reefs with linear transects: a study of storm damage. Journal of Experimental Marine Biology and Ecology 66:285-300. Rogers, C.S., L.N. McLain, and C.R. Tobias 1991. Effects of Hurricane Hugo (1989) on a coral reef in St. John, US VI. Marine Ecology Progress Series 78:1 89- 1 99. Sebens, K.P. 1994. Biodiversity on coral reefs: what are we losing and why? American Zoologist 34:115-133. Williams, A.H. 1979. Interference behavior and ecology of threespot damselfish {Eupomacentrus planifrons). Oecologia 38:223-230. 562 Table 1. Site information for AGRRA stony coral and algal surveys in the Virgin Islands. Site name Site Reef Latitude Longitude Survey Depth Benthic >25 cm % live stony code Type'/ (0 ' " N) (0'"W) Date(s) (m) transects stony corals coral cover Exposure m (#/10m) (mean ± sd) Northern Virgin Islands St John (shallow) Great Lameshur, 12 F/Windward 18 18,853 64 43.312 May 26 00 3 12 8.5 18.5 ±3.5 Donkey Great Lameshur, 13 F/Windward 18 19.094 64 43.390 Oct 23 99 5.5 25 4 13.5 ±6.0 VIERS Fish Bay East, II F/Windward 18 19.073 64 45.808 Nov 14 99 5 27 3.5 10.5 ±6.0 Inner Fish Bay West, 14 F/Windward 18 19.053 64 45.878 Aug 15 00, 5 12 8.5 7.0 ±6.0 Inner Oct 24 00 St. John (deep) Great Lameshur, 15 F/Windward 18 18.572 64 43.302 Aug 3 99, 11 10 10.5 48.5 ± 13.5 Tektite Nov 29 99 Great Lameshur, 16 F/Windward 18 18.831 64 43.596 July 23 99 13 26 4.5 16.0 ±8.0 Yawzi Fish Bay East, 9 F/Windward 18 18.948 64 45.777 Feb 08 00 6.5 23 4.5 12.0 ±5.0 Outer Eish Bay West, 10 F/Windward 18 18.850 64 45.845 Aug 02 99 7.5 29 4 13.0±6,5 Outer St. Thomas Brewer's Bay 2 F/Windward 18 20.670 64 59.157 May 26 99, 8.5 16 6.5 19.0 ±6.5 June 9 00 Buck Island 4 F/Windward 18 19.781 64 57.097 May 19 98, July 14 99 14 38 3.5 11.5 ±6.0 Caret Bay I F/Windward 18 22.421 64 59.371 May 21 98, July 17 00 9.5 17 6 23,5 ±8.5 Flat Cay 3 F/Windward 18 19.072 64 59.444 July 7 99, June 12 00 12 18 6 2L0±7.5 Sprat Bay 5 F-B/Windward 18 19.718 64 55.630 Aug 31 00 10 8 13 29.5 ±7.0 Virgin Gorda Eustatia Reef 18 F-B/Windward 18 30.50 64 20.250 July 24 00 8.5 11 9.5 17.0 ±6.0 Guana Is Iguana Head 17 F/Windward 18 28.477 64 34.941 Aug 06 99 10 13 8.5 30.0 ±9.5 St. Croix Cane Bay 6 F/Windward 17 46.230 64 48.530 Oct 14 99, Dec 16 99 9.5 13 7.5 30,5 ± 11.0 Long Reef 8 F-B/Windward 1746.118 64 41.490 Oct 13 99 13.5 13 7.5 16.0 ±3.5 Salt River East 7 F/Windward 17 47.227 64 45.330 Oct 13 99 10 17 6 14.0 ±4.0 Anegada Herman's Reef 19 B-B /Leeward 18 33.841 64 14.320 July 24 00 13 10 10 19.5 ±5.5 Horseshoe Reef 20 B-B /Windward 18 39.965 64 13.890 July 22 00 10.5 17 6 14.0 ±6.5 Jack Bay 21 F-B/Windward 18 44.961 64 19.246 July 22 00 9 23 4.5 8.0 ±4.5 W. Cow Wreck 22 F-B/Windward 18 45.164 64 24.596 July 23 00 8.5 14 7 1L5±3.5 ' F = fringing; F-B = fringing-barrier; B-B = bank-barrier 563 Table 2. Size and condition (mean ± standard deviation) of all stony corals (>25 cm diameter) by sites in the Virgin Islands. Site name Stony corals Partial-colony mortality (%) Stony corals (%) (Site code) # Diameter Recent Old Total Standing Bleached Diseased With fish (cm) dead bites Northern Virgin Islands St. John (shallow) Great Lameshur, Donkey (12) 101 95.0 ±46.0 2.5 ±3.0 53.5±31.0 56.0±3L0 2.0 25.0 0 25.5 Great Lameshur, VIERS (13) 104 53.0 ±27.5 4.5 ±7.0 28.5 ±30.5 33.0 ±30.0 0 38.5 6.0 39.5 Fish Bay East, Inner (11) 100 52.0 ±28.5 3.5 ±6.0 39.5 ±32.5 43.0±3L5 1.0 38.0 10.0 4.0 Fish Bay West, Inner (14) 102 55.0 ±36.0 3.0 ±5.0 50.5 ±37.5 53.0 ±36.5 4.0 12.5 2.0 0.0 St. John (deep) Great Lameshur, Tektite(15) 103 75.0 ±46.5 0.5 ±2.5 21.0± 18.5 21.5 ± 19.0 0 36.0 10.0 3.0 Great Lameshur, Yawyi ( 16) 112 47.0 ±25.0 1.0 ±3.5 21.5±24.0 22.5 ±25.0 0 21.5 5.5 2.5 Fish Bay East, Outer (9) 102 58.5 ±35.5 0.5 ± 1.5 44.5 ±37.0 45.0 ±37.0 1.0 10.0 6.0 4.0 Fish Bay West, Outer (10) 102 54.5 ±38.5 1.0±2.5 21.0±22.5 22.0 ±22.5 0 19.0 11.5 1.5 St. Thomas Brewer's Bay (2) 103 50.0 ±30.0 L5±3.0 37.0 ±33.5 38.5 ±34.0 0 8.0 9.5 9.5 Buck Island (4) 133 4L0±22.5 2.0 ±9.0 25.5 ±30.5 27.5 ±30.0 1.5 23.5 4.5 13.5 Caret Bay ( 1 ) 105 53.5 ±32.5 0.5 ±2.0 26.0 ±30.5 26.5 ±30.5 0 18.0 6.5 8.5 Flat Cay (3) 107 42.0 ±20.5 1.0 ±2.0 22.0 ±30.5 23.0 ±28.0 0 18.5 10.5 3.7 Sprat Bay (5) 105 52.5 ±30.5 3.5 ±5.0 28.5 ±24.0 32.5 ±23.0 0 15.0 6.0 7.5 Virgin Gorda Eustatia Reef (18) 103 56.0 ±42.0 2.5 ±7.0 30.0 ±29.0 32.5 ±29.5 0 24.0 5.0 15.5 Guana Is Iguana Head (17) 112 55.0 ±30.5 1.0 ±3.0 17.5 ±28.0 22.0 ±23.5 0 15.0 24.0 14.5 St. Croix Cane Bay (6) 98 63.0 ±41.0 L5±3.5 23.5 ±25.0 24.5 ±25.0 0 52.0 5.0 19.5 Long Reef (8) 100 48.5 ±26.0 0.5± 1.5 42.0 ±31.0 42.5 ±31.0 0 46.0 1.0 1.0 Salt River East (7) 103 37.5 ± 14.5 <0.5 ± 1.0 35.0 ±33.5 35.0 ±33.5 1.0 46.5 0 0 Anegada Herman's Reef (19) 100 54.5 ±32.0 2.0 ±4.0 40.5 ±30.0 42.5 ±29.5 3.0 25.0 1.0 25.0 Horseshoe Reef (20) 102 64.0 ±57.0 1.0 ±2.0 38.0 ±37.0 39.0 ±37.0 12.0 23.5 2.0 9.0 Jack Bay (21) 100 55.5 ±34.0 5.5 ±20.0 38.0 ±36.0 43.5 ±37.0 8.0 35.0 2.0 5.0 W. Cow Wreck (22) 100 46.5 ±25.0 1.0 ±2.0 44.5 ±33.5 45.5 ±33.5 0 28.0 3.0 9.0 564 Table 3. Algal characertistics, and density of stony coral recruits and of Diadema antillarum (mean ± standard deviation) by sites in the Virgin Islands. Site name # Macroalgae % relative % relative Macroalgal Recruits Diadema (Site code) quadrats % relative % absolute abundance abundanee Height Indices #/.0625m^ #/100m^ abundanee abundance turf algae crustose (cm) Relative coralline algae (Absolute)^ Northern Virgin Islands St. John (shallow) Great Lameshur, Donkey (12) 52 28.5 ±23.0 31.8 68.5 ±23.0 3.0 ±4.0 2.5± 1.5 71 (80) 5.0 ±0.5 26 Great Lameshur, VIERS (13) 50 23.0± 18.0 36.1 59.5±2L0 17.5 ± 14.0 2.0 ± 1.0 46 (72) 7.5 ± 1.0 5 Fish Bay East, Inner (11) 58 36.0 ±26.5 38.8 49.0 ±27.4 15.0 ± 17.0 3.0 ± 1.0 108(116) 3.5 ±0.5 7 Fish Bay West, Inner (14) 59 79.5 ±28.5 51.6 11.0±27.5 9.0 ± 15.5 3.5 ± 1.0 278 (181) 1.5 8 St. John (deep) Great Lameshur, Tektite (15) 54 76.0 ±30.0 44.6 17.5 ±23.5 6.5 ± 15.5 2.5 ± 1.5 190(112) 8.0 ± 1.0 0 Great Lameshur, Yawzi (16) 50 37.0 ±29.5 20.3 49.5 ±3 1.5 14.0 ± 13.0 1.5± 1.0 55(31) 10.0± 1.0 1 Fish Bay East, Outer (9) 50 6.0 ±7.0 6.8 75.5 ± 18.0 18.5 ± 17.0 L0± 1.0 6(7) 17.5 ± 1.5 0 Fish Bay West, Outer (10) 60 51.5 ±34.0 30.4 42.5 ±35.0 6.0 ±9.5 1.0 ±0.5 52 (30) 3.5 ±0.5 0 St. Thomas Brewer's Bay (2) 50 33.0 ± 17.0 47.1 62.5 ± 18.5 4.5 ±7.5 2.5 ±2.0 83 (118) 12.5 ± 1.0 4 Buck Island (4) 72 69.0 ±25.5 38.2 19.0 ±22.0 12.0± 15.0 2.5 ± 1.0 173 (96) 8.5 ± 1.0 0 Caret Bay (1) 52 55.0 ±35.0 40.9 7.0 ±20.0 38.0 ±30.0 2.0 ± 1.0 110(82) 7.5 ± 1.0 0 Flat Cay (3) 50 46.5 ±28.0 26.3 37.0 ±30.0 16.5 ± 19.0 2.0 ± 1.5 93 (53) 10.5 ± 1.0 1 Sprat Bay (5) 8 70.5 ± 28.5 57.6 *1 29.5 ±28.5 3.5 ±2.0 247 (201) 2.0 4 Virgin Gorda Eustatia Reef (18) 52 58.0±3L5 43.2 * 42.0± 31.5 2.0 ± 1.5 116(86) 3.0 ±0.5 6 Guana Is Iguana Mead (17) 50 8.0 ± 14.0 7.7 78.5 ±22.5 13.5 ±20.0 1.0 ± 1.0 8(8) 10.0 ± 1.0 0 St. Croix Cane Bay (6) 50 17.5 ± 17.0 35.2 67.0 ±20.0 15.0 ± 19.5 2.0± 1.0 35 (70) 6.0 ± 1.0 4 Long Reef (8) 70 25.0±2L5 18.9 66.0 ±24.6 8.5 ± 14.5 1.0 ±0.5 25(19) 16.0 ±2.0 0 Salt River East (7) 50 5.5 ± 10.5 3.2 89.0 ± 15.5 6.0 ±9.5 0.5 ±<0.5 3(2) 9.0 ± 1.0 0 Anegada Herman's Reef (19) 51 52.0 ±33.0 27.3 * 48.0 ±33.0 2.0 ± 1.5 104 (55) 4.5 ±0.5 0 Horseshoe Reef (20) 50 73.5 ±27.5 35.1 ♦ 26.5 ±27.5 2.0± 1.5 147 (70) 10.0 ±2.0 0 Jack Bay (21) 70 40.5 ±32.8 30.8 * 59.5 ±33.0 2.5 ±2.0 101 (77) 4.5 ±0.5 0 W. Cow Wreck (22) 53 65.7 ±28.0 42.8 * 34.3 ±28.0 3.5 ±4.5 230 (150) 7.5 ± 1.0 0 '* = not recorded ^Macroalgal index = % relative (or absolute) macroalgal abundance x macroalgal height. Appendix A. Qualitative assessment of human and natural impacts at 22 sites in the Virgin Islands. Site name Sediment Divers Fishing Exposure to Historical Historical (Site code) prevailing seas hurricane damage disease epizootics Northern Virgin Islands St. John (shallow) Great Lameshur, low low low low moderate moderate Donkey (12) Great Lameshur, low moderate Moderate? low moderate moderate VIERS (13) Fish Bay East, moderate low low low low moderate Inner (1 1) Fish Bay West, high low low low low moderate Inner (14) St. John (deep) Great Lameshur, low moderate moderate? low moderate high Tektite (15) (white plague) Great Lameshur, moderate moderate moderate? low high moderate Yawzi (16) (white plague) Fish Bay East, moderate low low low high moderate? Outer (9) Fish Bay West, high low low low high moderate? Outer (10) St. Thomas Brewer's Bay (2) moderate low moderate low low low Buck Island (4) low high low low low low Caret Bay ( 1 ) low low low moderate low low Flat Cay (3) low high moderate low low low Sprat Bay (5) low moderate moderate low low low Virgin Gorda Eustatia Reef low? high moderate? moderate moderate? moderate? (18) (white band) Guana Is Iguana Head (17) low? low? moderate? low low? low? St. Croix Cane Bay (6) moderate high low moderate low low? Long Reef (8) moderate high high moderate moderate high (white band) Salt River East (7) moderate high moderate moderate low low? Anegada Herman's Reef low low low? moderate high high (19) (white band) Horseshoe Reef low low low? high high high (20) (white band) Jack Bay (21) low moderate? moderate? high high high (white band) W. Cow Wreck low low low? high high high (22) (white band) 566 Figure 1. AGRRA survey sites in the Virgin Islands. (1) Caret Bay, (2) Brewer’s Bay, (3) Flat Cay, (4) Buck Island, (5) Sprat Bay, (6) Cane Bay, (7) Salt River, (8) Long Reef, (9) Fish Bay east outer, (10) Fish Bay west outer, (11) Fish Bay east inner, (12) Great Lameshur Donkey, (13) Great Lameshur VIERS, (14) Fish Bay west inner, (15) Great Lameshur Tektite, (16) Great Lameshur Yawzi, (17) Iguana head, (18) Eustatia Reef, (19) Herman’s Reef, (20) Horseshoe Reef, (21) Jack Bay, (22) West Cow Wreck. A RAPID ASSESSMENT OF CORAL REEFS IN THE VIRGIN ISLANDS (PART 2: FISHES) BY RICHARD S. NEMETH,* LESLIE D. WHAYLEN,^ and CHRISTY V. PATTENGILL-SEMMENS^ ABSTRACT The Atlantic and Gulf Rapid Reef Assessment (AGRRA) fish survey was conducted in depths of 3-14 m at 22 coral reef sites in the U.S. Virgin Islands (St. Croix, St. John, St. Thomas) and the British Virgin Islands (Anegada, Guana Island, Virgin Gorda). Total fish species richness, determined by using the roving diver technique was 72-160 species. A total of 14,441 fish from 54 species of selected families were counted in belt transects. The select fish fauna was dominated by the Scaridae (35%), Acanthuridae (35%) and Haemulidae (16%). Key herbivorous fish densities among sites ranged from ~4-49 fish/ 100m for scarids and from 1-60 fish/ 100m for acanthurids. Predatory fish densities among sites ranged from 0.2-5. 5 fish/lOOm for serranids and from 0-1 1 .5 fish/lOOm^ for lutjanids. The density of large herbivorous fishes showed a significant negative relationship with macroalgal index and with the percent of coral colonies occupied by territorial damselfish. Also significant were the positive relationships between total fish species richness and coral cover, between chaetodontid density and total fish species richness, and between pomacanthid density and depth. INTRODUCTION The Virgin Islands (18°20’ N, 64°50’W) lie between two major island archipelagoes: the Greater Antilles to the west and the Lesser Antilles to the southeast. With the exception of St. Croix, the northern United States Virgin Islands (USVI), together with British Virgin Islands (BVI) and the islands of Puerto Rico, rise from a geological shelf that is surrounded by deep water (Dammann, 1 969, unpublished report). This shelf covers approximately 3,200 km^, contains about 500 km of shelf edge and is generally less than 100 m in depth. St. Croix sits on a similar but smaller shelf that is separated from the northerly Virgin Islands by the 4,685 m-deep and 60 km-wide Virgin Islands trough. The three main U.S. islands of St. Thomas, St. John and St. Croix are Center for Marine and Environmental Studies, University of the Virgin Islands, MacLean Marine Science Center, 2 John Brewer's Bay, St. Thomas, UVSI 00802-9990. Email: memeth@uvi.edu 2 Reef Environmental Education Foundation, P.O. Box 246, Key Largo, Florida 33037. Email: leslie@reef.org, christy@reef org 568 surrounded by more than 90 uninhabited islands, cays, and rocks, which provide hard substrata suitable for the growth of coral reefs (Towle, 1970, unpublished report; Dammann and Nellis, 1992). Many of these reefs are adjacent to mangrove forests, seagrass beds and algal plains, which are important nursery habitats for reef-associated fishes (Austin, 1971; Parrish, 1989; Heck and Weinstein, 1989). The US VI, like other populated regions of the world, have witnessed steady declines in catches of commercially important marine fishes (Caribbean Fisheries Management Council, 1985, unpublished report; Roberts, 1997). Millions of people in developing countries depend upon coral reef fishes for their livelihoods (Roberts, 1993). However, declining fisheries cannot only be attributed to overfishing but to a suite of other factors since evidence suggests that degradation of coral reefs can influence reef fish communities (Bell and Galzin, 1984; Bouchon-Navaro et ah, 1985). Over the past 20 years, eight major hurricanes, numerous disease outbreaks, sporadic bleaching events, ship groundings, anchor damage, plus inputs of sediment and pollution have caused excessive damage to the coral reefs surrounding the Virgin Islands (Gladfelter, 1982; Rogers et ah, 1991; Rogers and Garrison, 2001). The US VI fishing community has remained relatively stable since the early 1900’s. Licensed fishermen have consistently numbered approximately 400 between 1930 and today, even though the total population of the islands has increased from 22,012 to about 100,000 (Fielder and Jarvis, 1932; Department of Commerce Census, 2000). Since the early 1970’s, fish stocks in the near shore US VI have been declining steadily despite government- funded research and development programs for the deep-water fisheries around the Virgin Islands plateau (Brownell and Rainey, 1971, unpublished report). The catch of fishery products fluctuated between 550,000 and 1,650,000 pounds per year between 1930 and 1997 (Fielder and Jarvis, 1932; de Graff and Moore, 1987, unpublished report; Tobias, 1997, unpublished report), but the fishing effort increased dramatically. For example, in the USVI the average number of traps fished per full-time fisherman increased from 4 in 1930, to 8 in 1967, to >100 in 1997 (Fielder and Jarvis, 1932; Dammann, 1969; unpublished report; T. Tobias, USVI Fish and Wildlife, 1998). The maximum number of traps fished by a single fisherman in 1930 was 30 traps whereas today it can be as high as 2000 traps (Fielder and Jarvis, 1932; Downs and Petterson, 1997, unpublished report). The pursuit of declining stocks not only puts fish populations at risk of commercial extinction but often increases reef damage associated with fishing (Sladek Nowlis and Roberts, 1998; Appeldoom, 2001, unpublished manuscript; Nemeth, unpublished data). Increased fishing pressure on nearshore and offshore fish populations has already resulted in the collapse of several species. For example, in the 1970’s unregulated fishing in spawning aggregations of Nassau grouper (Epinephelus striatus) and yellowfm grouper {Mycteroperca venenosd) off St. Thomas led to the commercial extinction of these species (Olsen and LaPlace, 1978; Beets and Friedlander, 1992, 1999). In the absence of Nassau grouper, fishermen targeted red hind {E. guttatus) spawning aggregations. Within 10 years, the St. Thomas red hind population was on the verge of collapse when a seasonal (December through February) closed area at the red hind spawning aggregation site was implemented in 1990 (Beets and Friedlander, 1992). Within 6 years of the closure. Beets and Friedlander (1999) reported that an increase in 569 their average size and an improved sex ratio (red hind are protogynous hermaphrodites) were indications of red hind population recovery. Without proper management tools, sustained fishing mortality can decrease the capacity, productivity and genetic diversity of the fishery (Bohnsack, 1990). Moreover, selective removal of ecologically important species can lead to ecosystem damage and produce major ecological shifts (Hay, 1984; Roberts, 1995; Botsford et al., 1997). In response to these threats to coral reefs, Ginsburg et al. (1996) initiated a process of rapid reef assessment. After development of the AGRRA protocols, a Caribbean-wide effort to assess the condition of coral reefs and associated fishes throughout the region was launched. The Center for Marine and Environmental Studies of the University of the Virgin Islands (UVI) joined the effort and set out to conduct a broad survey of the fish assemblages of the Virgin Islands with a focus on commercially and ecologically important species using the AGRRA protocol. This paper reports on the initial findings of our assessment. Results of the benthic surveys are given in Nemeth et al. (this volume). METHODS An assessment of fish species abundance and diversity was conducted at 22 sites in the Virgin Islands (Fig. 1) between May 1998 and July 2000. Site selection criteria in the USVI and the BVI varied among the different islands, but most (18/22) choices were made for strategic reasons. In St. Thomas, five sites were selected based on their inclusion in long-term sedimentation and reef fish monitoring projects, their proximity to the UVI MacLean Marine Science Center and the presence of historical data (i.e., Rogers, 1982, unpublished report; Nemeth and Sladek Nowlis, 2001). Three popular recreational diving sites were selected in St. Croix, one of which is within the Salt River Bay National Historic Park and Ecological Preserve. The eight sites off St. John were part of a study comparing sedimentation rates between Great Lameshur Bay, which is within the National Park Boundary, and Fish Bay, which has been experiencing heavy development within its watershed. Four of these sites were shallow reefs (< 6 m) located inside the bays and four sites were deeper reefs (> 6 m) located outside the bays. Four of the six sites surveyed in the BVI (Iguana Head on Guana Island, Eustatia Reef on Virgin Gorda, West Cow Wreck Bay and Herman’s Reef on Anegada) were selected haphazardly. Eustatia Reef, off Virgin Gorda, is a heavily visited dive site whereas Guana Island is a private island with few inhabitants. The other two sites on Anegada (Jack Bay and Horseshoe Reef-a designated protected area) were selected because of historical surveys (Dunne and Brown, 1976; West Indies Laboratory, 1983, unpublished reports). The reefs of Anegada were included in the AGRRA survey to provide a remote reference site with low human population and little landmass. The major difference in fishing regulations between the USVI and the BVI is that spearfishing is prohibited in the BVI. A qualitative assessment of natural and human impacts, including fishing pressure, at each site surveyed in the Virgin Islands is given in Appendix A of Nemeth et al. (this paper). The AGRRA fish protocol Version 2.0 (see Appendix One, this volume) was used except off St. Croix where Version 2.1 was employed. Three divers (authors) constituted the primary fish survey team with R. Nemeth conducting the majority of surveys on all islands except those off St. Croix. Fish identification guides included Humann (1994), 570 Stokes (1984) and Robins and Ray (1986). Visual fish counts were conducted along at least ten 30 x 2 m transects, and one 30-minute roving diver survey was conducted at each site except in Brewer’s Bay, St. Thomas (n = 2 surveys) and off St. Croix, where longer (60-90 minute) survey were conducted in Cane Bay (n = 3), Long Reef (n = 5) and Salt River (n = 6). During fish transects, transect width and fish lengths (cm) were estimated using a 1 m-wide T-bar constructed of polyvinylchloride pipe marked in 5 cm and 10 cm increments. In all sites except St. Croix, scarids (parrotfish) and haemulids (grunts) less than 5 cm in length were counted and identified to species whenever possible. Using transects as replicates, the average density (#/100m ) and size (cm) of each species and family were calculated for each site and island group (see below). The size distributions of herbivores [parrotfish, surgeonfish (acanthurids) and the yellowtail damselfish, Microspathodon chrysurus)] and of the AGRRA carnivores [select groupers (serranids) and all lutjanids] were also calculated. Data were summarized by island groups within three geographic areas to examine general trends in reef fish assemblages and to be comparable with the Virgin Islands benthic data which are presented in Nemeth et al. (this volume). The geographic areas were: (1) Anegada; (2) the shallow and deep reefs of the northern US VI and BVI (NVI); and (3) St. Croix. Anegada was considered a geographic unit because of its isolation from the other Virgin Islands and its unique geology (low coral island). St. Croix was considered a geographic unit also because of its isolation from the NVI, its unique geology (sedimentary/carbonate) and because it is completely within the Caribbean Sea. The NVI (USVI = St. John, St. Thomas; BVI = Guana Island, Virgin Gorda) were grouped because of their close proximity to one another, their similar geologic origins and topographies (high volcanic islands) and their exposure to both Atlantic waters from the north and Caribbean waters from the south. The shallow sites off St. John were analyzed separately from the deep reefs within the NVI archipelago. RESULTS A total of 256 fish transects and 34 roving diver surveys were conducted at 22 sites throughout the USVI and the BVI (Table 1). During visual belt transects of the selected AGRRA fishes (Appendix One), divers recorded 54 species and counted a total of 8,227 fish (n= 14,441 fish when grunts and parrotfishes <5 cm in Anegada and the NVI are included). Fish species richness, based upon the maximum number of species seen during any single roving diver survey, varied between a high of 74 species in Cane Bay (St. Croix) and a low of 40 species in Fish Bay (shallow St. John) with >90% (20/22) having values between 40 and 70 species (Table 1). The average number of fish species by island group was greatest in St. Croix (68 species), followed by the deep NVI reefs (59), Anegada (52) and the shallow St. John sites (47), although these differences were not significant (F3,29 ^ 2.75, p = 0.061). Total cumulative reef fish richness by island group was 160 species in the NVI (n = 12 surveys at 1 1 sites), 131 species in St. Croix (n = 14 surveys at 3 sites), 98 species in Anegada (n = 4 surveys at 4 sites) and 72 species in the shallow St. John sites (n = 4 surveys at 4 sites). The additional roving diver surveys conducted in St. Croix and the large number of deeper sites in the NVI most likely contributed to a higher number of fish species seen within these island groups. To 571 illustrate, divers in St. Croix eounted: 58, 68 and 74 fish species during three roving surveys in Cane Bay; 45, 53, 63, 65, and 68 species during five roving surveys in Long Reef; and 37, 49, 54, 58, 60 and 63 species during six roving surveys in Salt River. The cumulative numbers of fish species seen in Cane Bay, Long Reef and Salt River reefs during these multiple surveys were 92, 88 and 117, respectively (Table 1). Based on an analysis of the six roving diver surveys at Salt River where search times averaged 73 minutes per survey, a species accumulation curve showed that at least four such surveys (i.e. 4.8 hours search time) are necessary to record at least 90% of the fish species and at least six surveys (i.e., 7.3 hours search time) to approximate total species diversity at a site. Thus, depending upon the skill level of the diver, species diversity estimates based on a single 60-minute roving diver survey may represent only 40% to 70% of the actual number of species present in a reef The 25 most commonly seen species during roving diver surveys were represented by 1 1 fish families (Pomacentridae, Scaridae, Labridae, Serranidae, Acanthuridae, Holocentridae, Haemulidae, Chaetodontidae, Gobiidae, Mullidae, and Tetraodontidae) with Thalassoma bifasciatum, Acanthurus coeruleus, Sparisoma aurofrenatum, S. viride and Haemulon flavolineatum being ubiquitous in all sites (Table 2). Of the 54 AGRRA species sighted in belt transects, scarids and haemulids were each represented by 9 species, lutjanids by 7 species, serranids and pomacanthids each by 6 species, balistids by 5 species, chaetodontids by 4 species, acanthurids by 3 species, labrids by 2 species, and pomacentrids, sphyraenids and carangids each by 1 species. The 25 most abundant of these species (Table 3) represented nearly 94% of all fish sighted in the belt transects. Numerically the AGRRA fish fauna in the Virgin Islands was dominated by the families Scaridae (35%, all >5 cm), Acanthuridae (35%) and Haemulidae (16%, all >5 cm) (Fig. 2, Table 3). The herbivorous scarids and acanthurids represented 7 of the 10 most abundant species, with the ocean surgeonfish (Acanthurus bahianus) ranked first overall. Three carnivores, the French grunt {Haemulon flavolineatum), tomtate {H. aurolineatum) and yellowtail snapper (Ocyurus chrysurus) were also ranked among the top 10 most abundant species (Table 3). Although newly recruited grunts were not included in this analysis, high densities of juvenile haemulids (<5 cm) were observed in the NVI off Guana Island (710.7 grunts/ 100 m ) and Buck Island, St. Thomas (10.7 grunts/100 m^), as well as in VIERS (8.3 grunts/ lOOm^) and Fish Bay East Inner (3 grunts/ 100 m^) in shallow St. John. High densities of juvenile scarids were also recorded. Scarids <5 cm comprised 24-77% of parrotfish densities in Anegada and the NVI (Fig. 3). We found a general increase in the density of juvenile scarids off Anegada (the easternmost reefs) through the lower BVI to the shallow St. John sites where they were most abundant (Fig. 3). Juvenile densities dropped in the deep St. John sites and were slightly higher around St. Thomas. A weak negative relationship between the density of juvenile scarids <5cm and depth (r^ = 0.18, F2o,i = 4.48, p< 0.05) was found in a regression analysis. Herbivorous fish densities (# fish/ 100 m^) among sites ranged from 3.8 (Fish Bay West Inner, St. John) to 48.8 (Great Lameshur, Yawzi, St. John) for the Scaridae and 1.0 (Brewer’s Bay, St. Thomas) to 60.5 (Fish Bay East Outer, St. John) for the Acanthuridae (Table 4). Overall, the abundance of parrotfishes (>5 cm) and surgeonfishes were similar in both shallow (11.6 scarids versus 13.1 acanthurids/ 100 m ) and deeper (19.4 scarids versus 18.2 acanthurids/ 100 m ) reefs. Predatory fish densities (# fish/ 100 m ) among 572 Figure 2. Fish density (mean no. fish/100 ± standard error) for AGRRA fishes in shallow St. John, deeper NVI (St. Thomas, St. John >6 m deep. Guana Is., Virgin Gorda), St. Croix and Anegada. Other = Bodianus rufus, Caranx ruber, Lachnolaimus maximus, Microspathodon chrysurus, Sphyraena barracuda. sites ranged from 0.2 (Fish Bay East Inner, St. John) to 5.5 (Cane Bay, St. Croix) for the Serranidae, and from 0.0 (Herman’s Reef, Anegada) to 1 1.5 (Sprat Bay, St. Thomas) for the Lutjanidae (Table 4). It is unusual to note that no snappers were seen at Herman’s Reef during transects or roving diver surveys and no grunts were counted during transects at Horseshoe Reef (one French grunt was seen during a roving diver survey). The size distributions of fishes comprising the two major feeding guilds (herbivores and carnivores) were found to be fairly similar among the deeper NVI reefs and those of St. Croix (Fig. 4). Anegada was unique in that it had twice the density of 21- 30 cm carnivores and 1 1-20 cm herbivores as the other sites and an abundance of juvenile acanthurids <5 cm (Fig. 4). Jack Bay was unusual in that it had considerably smaller groupers than other sites in Anegada (Table 5). The shallow St. John sites had 573 Figure 3. Density (mean no. fish/ 100 m^) of scarids <5cm and >5 cm in length for Anegada (four reefs), Lower BVI (two reefs), shallow St. John (four reefs), deeper St. John (four reefs) and St. Thomas (five reefs). Scarids <5 cm were not recorded in St. Croix. Figure 4. Size composition (mean density ± standard error) of (A) carnivores (lutjanids, large serranids) and (B) herbivores (acanthurids, scarids >5 cm, Microspathodon chrysurus) in shallow St. John, deeper NVI, St. Croix and Anegada. 574 substantially fewer groupers and snappers in the 21-30 cm size class (Fig. 4) but over twice as many newly settled (<5 cm) scarids as the other sites (Fig. 3). Within the deeper NVI sites, two reefs in the BVI (Virgin Gorda and Guana Island) and most of the deeper sites around St. John had larger groupers than the reefs around St. Thomas (Table 5). With the exception of Buck Island and Sprat Bay, the reefs off St. Thomas also tended to have smaller snappers and parrotfish relative to the other NVI sites. One site in Great Lameshur Bay (VIERS) had smaller groupers than the other shallow sites off St. John. Similarly, Cane Bay generally had smaller snappers than the other two reefs off St. Croix (Table 5). A series of multiple regression analyses were conducted between the average fish density or the total fish species richness and each of the following: depth, live stony coral cover and, for “large” (>25 cm diameter) stony corals, mean size and recent partial- colony mortality (Nemeth et ah, this volume). Only one of these comparisons was significant. Total fish species richness (using results from a single roving diver survey per site) showed a positive relationship with percent live stony coral cover (Fig 5). Upon further analysis of fish guilds or individual species, some other interesting patterns emerged. A significant negative relationship was found between the density of herbivorous fishes greater than 5 cm in length and macroalgal index calculated from percent absolute abundance of macroalgae (Fig 6). This relationship remained significant even when herbivore density was compared to macroalgal index calculated with percent relative abundance of macroalgae (r^ = 0.28, p < 0.04) or when compared directly to absolute macroalgal abundance (r^ = 0.64, p < 0.001). Significant positive relationships were also found between the density of butterfly fish (Chaetodontidae) and total fish species richness from either a single roving diver survey per site (r = 0.19, p < 0.04) or from results of multiple roving diver surveys (Fig. 7). However, no significant relationship was found between the density of chaetodontids and percent live stony coral cover (p < 0.62). A significant positive relationship was found between the density of angelfish (Pomacanthidae) and depth (Fig. 8). Finally an interesting negative relationship was found between the density of large herbivores (>10 cm) and the percent of large stony corals occupied by territorial damselfishes (Stegastes planifrons, S. diencaeus and, S.fuscus; Fig. 9). Figure 5. Regression between total fish species richness and percent live stony coral cover (r^ = 0.30, p < 0.01), by site in the Virgin Islands. 575 Figure 6. Regression of herbivore (acanthurids, scarids >5 cm, Microspathodon chrysurus) density versus the macroalgal index (absolute abundance of macroalgae x macroalgal height) by site in the Virgin Islands, r^ = 0.24, F2o,i = 6.33, p < 0.02. Figure 7. Regressions between chaetodontid density and total fish-species richness (r^ = 0.564, < 0.0001) by site in the Virgin Islands. 576 Figure 8. Regression of pomacanthid density and depth (r^ = 0.32, p < 0.005) by site in the Virgin Islands. % large stony corals occupied by territorial damselfish Figure 9. Relationship between the density of adult (>10 cm length) herbivores (acanthurids, scarids, Microspathodon chrysurus) and the percent of large (>25 cm) stony corals occupied by territorial damselfish by site in the Virgin Islands. Nonlinear stepwise regression analysis produced a significant fit (r‘ = 0.97, p < 0.0001) with an inverse second order quadratic equation (F = yO +a/x = b/x^). DISCUSSION Fish assemblage structure was relatively similar among island groups within the USVI and BVI. Herbivorous fishes, especially the parrotfishes and surgeonfishes, dominated the reefs at all sites and, overall (>5 cm scarids only), comprised 72% of the 577 fish fauna. The shallow St. John sites deviated from this pattern with the >5 cm Haemulidae showing a five-fold increase in abundance relative to the deeper reef areas throughout the region. The abundance of haemulids in shallow reefs can be attributed most likely to the proximity of seagrass beds (Parrish, 1989; Muehlstein and Beets, 1992; Nagelkerken et ak, 2000a,b,c, 2001), which are prevalent along the south coast of St. John. Juvenile scarids were a major component (24% to 77%) of parrotfish densities in Anegada and the NVI. There was a general increase in density of juvenile parrotfish from the easternmost reefs of Anegada to the shallow reefs of St. John, which may indicate an upstream source of larval fish recruitment (Roberts, 1997). However, a weak negative relationship with depth or the lower number of large carnivores on the shallow reefs of St. John may also explain some of the spatial variation in juvenile scarid densities. Although temporal variation in recruitment patterns may have influenced juvenile scarid densities on the shallow St. John reefs, many of the fish surveys were conducted during similar time periods. The remaining selected fish families [i.e., snappers, groupers, angelfish, butterflyfish, and triggerfish (balistids)] were similar in abundance among the deep reef areas but much less common in the shallow St. John sites. Total fish species richness was also similar among the island groups, although estimates were greatly influenced by the number of sites visited or the number of roving diver surveys performed at a site. Based on species accumulation curves, a minimum of six roving diving surveys or more than seven hours search time would be needed to approximate the actual species diversity at a site. Although this level of effort would not be practical considering that AGRRA is supposed be a rapid assessment technique, it is recommended that future AGRRA surveys conduct at least one, but preferably two, 60-minute roving diver surveys. Multiple regression analyses detected some interesting regional patterns of abundance associated with benthic parameters which accounted for a significant portion of the variability in total fish species richness and in the density of butterflyfish and angelfish. The significant relationship between total fish species richness and percent cover of live stony corals (Fig. 5) suggests that species packing occurs where well- developed reefs provide greater structural complexity (Risk, 1972; Luckhurst and Luckhurst, 1978; Bell and Galzin, 1984; Roberts and Ormond, 1987). In earlier studies of chaetodontid fish assemblages around the world, significant relationships were found between their densities and the percent living coral cover and coral diversity (Bell et ak, 1985; Bouchon-Navaro et ak, 1985; Bouchon-Navaro and Bouchon, 1989; Roberts et ak, 1992). Our analysis differed from these results in that the density of butterflyfishes was not associated with any benthic factor. A positive relationship between butterflyfish density and total fish species richness in the Virgin Islands (Fig. 7) complemented the results of Roberts et ak (1992), who found a positive relationship between the density and total abundance of butterflyfishes and species richness in the Red Sea. Important differences in fish size distribution patterns were apparent among the island groups. Within the Virgin Islands archipelago, Anegada is the most remote and least populated island, and its reefs probably experience the lowest level of fishing pressure relative to the other islands. Moreover, large portions of the extensive reef system surrounding Anegada, along with 12 other sites in the BVI, are marine conservation parks protected by the BVI National Parks Trust. All commercial and recreational fishing is prohibited within marine park boundaries (BVI National Parks 578 Trust, unpublished). These factors probably contributed to a greater abundance of the large-size classes of predators and herbivores in Anegada as well as in the two other BVI reefs (Iguana Head, Guana Island and Eustatia Reef, Virgin Gorda). Patterns also emerged when the average sizes of several commercially important fish families were compared among islands. In general, most sites off St. Thomas had smaller groupers, snappers, grunts and parrotfishes than deeper reefs off other islands (i.e., St. John, St. Croix, Virgin Gorda and Guana Island). These differences were probably due to a suite of factors specific to each island. For example, St. Thomas has the greatest human population density of all the islands, the largest number of seasonal workers from other Caribbean islands (who often supplement their diet by fishing in near shore waters), and the highest number of tourists who participate in recreational or sport fishing activities (US Department of Commerce, 2000; Virgin Islands Department of Fish and Wildlife, 2001, unpublished report). Fishing regulations, primary target species, and gear types may also influence harvest rates. In a survey of recreational fishers in the Virgin Islands, Jennings (1992) reported that nearly 65% of the fishers targeted groupers and snappers. He also found that up to 8% speared fish in the St. Thomas/St. John area, whereas none spear fished in St. Croix. In contrast, spear fishing is prohibited throughout the BVI and within the Virgin Islands National Park waters surrounding approximately 50% of St. John. In the Florida Keys, sites protected from spear fishing had higher densities and larger sizes of several species especially snappers and grunts (Roberts and Polunin, 1993). Even low levels of protection within an area have resulted in increased biomass of commercially important fishes (Bohnsack, 1996; Roberts and Hawkins, 1997). In a recreational fisher survey conducted in the US VI in 2001 by the Virgin Islands Division of Fish and Wildlife, 67% of the people interviewed indicated that they spear fished or bottom fished less than 3 miles from shore (unpublished report). These and other factors would greatly intensify the fishing pressure on near-shore stocks and result in over-harvest around the St. Thomas area (Jennings, 1992). By focusing on the variation in fish size within each island or island group, we also observed that some sites had consistently smaller fish across most commercially important families (Table 5). These sites included the VIERS shallow reef and Tektite deep reef (St. John), Brewer’s Bay (St. Thomas), Cane Bay (St. Croix ) and Jack Bay (Anegada). Although many natural factors may contribute to the impacts on reef fisheries, one of the most important similarities among all of these sites seems to be accessibility to skin or scuba divers. Easy access from shore and popularity of a dive site (i.e., presence of a public mooring) will concentrate fishing activity in these locations. This is especially true when commercial dive charters allow patrons to spear fish. VIERS reef. Brewer’s Bay, Cane Bay and Jack Bay are unique among all the sites surveyed since they are easily accessible from shore by skin and scuba divers who may participate in spear fishing (R. Nemeth, personal observation). Brewer’s Bay, in particular, is a beach frequently (e.g. weekly) used by local spear fishermen (R. Nemeth, personal observation). Tektite reef is also a popular dive destination, and may experience increased fishing pressure even though it is within the Virgin Islands National Park. The fact that Rogers and Beets (2001) found no apparent difference in fish assemblage structure inside or outside the national park boundaries indicates that fishing pressure is still substantial at some sites within the park. 579 Further support of the hypothesis that size structure is largely determined by access to divers can be gained by looking at reefs off islands that consistently had larger fish in most commercially important families. Sites such as Caret Bay (St. Thomas) and Fish Bay East Outer (St. John) are rarely, if ever, visited by skin or scuba divers, and Salt River (St. Croix), although a popular dive area, is within the Salt River Bay National Historic Park and Ecological Preserve. Thus fishing pressure at these sites is very low relative to more popular or unprotected sites. Overall densities of parrotfishes and surgeonfishes in the shallow and deeper reefs of the Virgin Islands were comparable to those reported by Lewis and Wainwright (1985) with the exception that parrotfishes of all sizes in the shallow St. John sites were 7.5 time more abundant than in Belize (50.9 versus. 6.7 scarids/lOOm^). The significant negative relationship between herbivore density and macroalgal index (Fig. 6) suggests that grazing pressure of herbivorous fishes is an important factor in determining the structure of tropical benthic communities (Lewis and Wainwright, 1985; Lewis, 1985, 1986). Grazing pressure was strongly correlated to herbivore density in Belize (Lewis, 1985, 1986). Lewis (1986) showed a nearly 30% increase in the cover of macroalgae (primarily Padina and Dictyota) relative to controls when herbivorous fishes were excluded from experimental plots in back-reef habitats in Belize. Total macroalgal index calculated from Lewis (1986) was 150 within shallow back-reef exclusion cages compared to an average macroalgal index of 126 on shallow reefs and 99 on deeper reefs of the Virgin Islands archipelago (Nemeth et ah, this volume). In the Virgin Islands, the density of large (>10 cm) herbivores (parrotfishes and surgeonfishes) was negatively related to the percent of large stony corals occupied by territorial damselfishes (Fig. 9). The presence of territorial damselfishes, which aggressively defend their algal turfs, has been shown to have a dramatic effect on the social behavior, abundance, and feeding strategies of other herbivorous competitors (Williams, 1979; Hourigan, 1986). In the Caribbean and Pacific, several studies have shown that parrotfishes and surgeonfishes form large schools to feed in the territories of other species, especially territorial damselfishes (Vine, 1974; Robertson et al., 1976; Hourigan, 1986). Robertson et al. (1976) found that schooling parrotfishes and surgeonfishes, which settled and fed en masse within damselfish territories, had higher feeding rates and fewer aggressive episodes than non-schooling individuals. The experimental removal of territorial damselfish within a defined area typically results in a rapid increase in the abundance of other herbivorous fishes (Robertson et al., 1976; Hourigan, 1986; Nemeth, unpublished data). The negative relationship between the density of large herbivores and the density of territorial damselfish could be due to an actual competitive exclusion of herbivores in the area and/or an artifact of the belt- transect method, which may underestimate the densities of schooling herbivores. By comparing the relative densities of herbivores from fish transects and roving diver surveys, we found that density estimates were comparable in magnitude at 85% (1 1/13) of the sites where damselfish occupied less than 10% of the coral heads but in only one of the six (17%) sites where damselfish occupied more than 10% of the coral heads. This suggests that, in areas with high densities of territorial damselfishes, the abundance of large herbivores were underestimated using fish transects since roving schools of parrotfishes and surgeonfishes can move unpredictably over large areas of coral reef (Robertson et al. 1976). 580 Considering the wide range of benthic conditions, human factors, and management regulations that could potentially influence fish assemblage structures, we were surprised to find that the abundance and size structure of commercially and ecologically important species were relatively similar among island groups of the Virgin Islands. Within the island groups, the primary notable differences were the larger size structure of herbivores and carnivores in Anegada and the greater abundance of haemulids in the shallow reefs of St. John. The significant relationships between selected fish families and various benthic parameters generally supported the observations of previous studies within the Caribbean and other tropical regions. Within each island, site- specific differences in fish-size structure may be explained by variations in fishing pressure from recreational fishermen, including spear fishers. ACKNOWLEDGMENTS We thank the AGRRA Organizing Committee which facilitated financial support through the United States Department of Interior. Additional surveys around St. John and St. Thomas were made possible through additional funding for related projects from the US Geological Survey via the University of the Virgin Islands Water Resources Research Institute and from the Puerto Rico Sea Grant College Program. Many thanks to Guana Island Resort and Lianna Jarecki for supporting the UVI assessment team while at Guana Island. This large-scale fish assessment could not have been accomplished without our dedicated divers: Adam Quandt, Laurie Requa, and Reef Environmental Education Foundation (REEF) volunteer divers Clint Whitaker and Denise Mizell. Thanks also for the support of UVI’s dive officer and captain, Steve Prosterman. We are especially grateful to Judy Lang for all her excellent editorial comments. REFERENCES Austin, H.M. 1971. Survey of the ichthyofauna of the mangroves of western Puerto Rico during December, 1967 - August, 1968. Caribbean Journal of Science 1 1:27-39. Beets, J., and A. Friedlander 1 992. Stock analysis and management strategies for red hind, Epinephelus guttatus, in the U.S.V.I. Proceedings of the Gulf and Caribbean Fisheries Institute 42:66-79. Beets, J., and A. Friedlander 1999. Evaluation of a conservation strategy: a spawning aggregation closure for red hind, Epinephelus guttatus, in the U.S. Virgin Islands. Environmental Biology of Fishes 55:91-98. Bell, J.D., and R. Galzin 1984. Influence of live coral cover on coral reef fish communities. Marine Ecology Progress Series 15:265-274. 581 Bell, J., M. Harmelin-Vivien, and R. Galzin 1985. Large scale spatial variation in abundance of butterflyfishes (Chaetodontidae) on Polynesian reefs. Proceedings of the Fifth International Coral Reef Congress, Tahiti 5:421-426. Bohnsack, J.A. (ed.) 1990 The potential of marine fisheries reserves for reef management in the U.S. Southern Atlantic. Technical MemorandumF\MFS-^TFC-26\ Contr. No. CRD/89-90. 14 pp. Bohnsack, J.A. 1996. Maintenance and recovery of reef fishery productivity. Pp. 282-313. In: N.V.C. Polunin and C.M. Roberts (eds.). Reef Fisheries. Chapman and Hall, London. Botsford L.W., J.C. Castilla, and C.H. Peterson 1997. The management of fisheries and marine ecosystems. Science 277:509-514. Bouchon-Navaro, Y., C. Bouchon, and M.L. Harmelin-Vivien 1985. Impact of coral degradation on a chaetodontid fish assemblage (Moorea, French Polynesia). Proceedings of the Fifth International Coral Reef Congress, Tahiti 5:427-432. Bouchon-Navaro, Y., and C. Bouchon 1989. Correlations between chaetodontid fishes and coral communities of the Gulf of Aqaba (Red Sea). Environmental Biology of Fishes 25:47-60. Dammann, A.E., and D.W. Nellis 1992. A Natural History Atlas to the Cays of the US Virgin Islands. Pineapple Press, Inc. Sarasota, FI. 160 pp. Fielder, R.H., and N.D. Jarvis 1932. Fisheries of the Virgin Islands of the United States. U. S. Department of Commerce (Bureau of Fisheries). Investigational Report 14:1-32. Ginsburg, R.N., R.P.M. Bak, W.E. Kiene, E. Gischler, and V. Kosmynin 1996. Rapid assessment of reef condition using coral vitality. Reef Encounter 19:12-14. Gladfelter, W.G. 1982. White-band disease in Acropora palmata: implications for the structure and growth of shallow reefs. Bulletin of Marine Science 32:639-643. Hay, M.E. 1984. Patterns of fish and urchin grazing on Caribbean coral reefs: are previous results typical? Ecology 65:446-454. Heck, K.L., and M.P. Weinstein 1 989. Feeding habits of juvenile reef fishes associated with Panamanian seagrass meadows. Bulletin of Marine Science 45:629-636. Hourigan, T.F. 1986. An experimental removal of a territorial pomacentrid: effects on the occurrence and behavior of competitors. Environmental Biology of Fishes 15:161-169. Humann, P. 1994. Reef fish identification: Florida, Caribbean, Bahamas. 2^^ edition. New World Publications, Inc. Jacksonville, Florida. 396 pp. 582 Jennings, C.A. 1992. Survey of non-charter boat recreational fishing in the U.S. Virgin Islands. Bulletin of Marine Science 50:342-351. Lewis, S.M. 1985. Herbivory on coral reefs: algal susceptibility to herbivorous fishes. Oecologia 65:370-375. Lewis, S.M. 1986. The role of herbivorous fishes in the organization of a Caribbean reef community. Ecological Monographs 56:183-200. Lewis, S.M., and P.C. Wainwright 1985. Herbivore abundance and grazing intensity on a Caribbean coral reef Journal of Experimental Marine Biology and Ecology 87 :2 1 5-228. Luckhurst, B.E., and K. Luckhurst 1978. Analysis of the influence of substrate variables on coral reef fish communities. Marine Biology 49:317-323. Muehlstein, L.K., and J. Beets 1992 Seagrass declines and their impact on fisheries. Proceedings of the Gulf and Caribbean Fisheries Institute 42:55-65. Nagelkerken, L, M. Dorenbosch, W.C.E.P. Verbeck, E. Cocheret de la Moriniere, and G. van der Velde 2000a. Importance of mangroves, seagrass beds and the shallow coral reef as a nursery for important coral reef fishes, using a visual census technique. Estuarine, Coastal and Shelf Science 5 1 :3 1-44. Nagelkerken, L, M. Dorenbosch, W.C.E.P. Verbeck, E. Cocheret de la Moriniere, and G. van der Velde 2000b. Importance of shallow- water biotopes of a Caribbean bay for juvenile coral reef fishes: patterns in biotope association, community structure and spatial distribution. Marine Ecology Progress Series 202:175-192. Nagelkerken, L, M. Dorenbosch, W.C.E.P. Verbeck, E. Cocheret de la Moriniere, and G. van der Velde 2000c. Day-night shifts of fishes between shallow-water biotopes of a Caribbean bay, with emphasis on the nocturnal feeding of Haemulidae and Lutjanidae. Marine Ecology Progress Series 194:55-64. Nagelkerken, L, S. Kleijnen, T. Klop, R.A.C.J van den Brand, E.C. de la Moriniere, and G. van der Velde 2001 . Dependence of Caribbean reef fishes on mangroves and seagrass beds as nursery habitats: a comparison of fish faunas between bays with and without mangroves/seagrass beds. Marine Ecology Progress Series 214:225-235. Nemeth, R.S., and J. Sladek Nowlis 2001 . Monitoring the effects of land development on the near-shore reef environment of St. Thomas U.S. Virgin Islands. Bulletin of Marine Science 69:759-775. Olsen D.A., and J.A. LaPlace 1978. A study of a Virgin Islands grouper fishery based on a breeding aggregation. Proceedings of the Gulf and Caribbean Fisheries Institute 3 1 : 130-144. 583 Parrish, J.D. 1989. Fish communities of interacting shallow- water habitats in tropical oceanic regions. Marine Ecology Progress Series 58:143-160. Risk, M.J. 1972. Fish diversity on a coral reef in the Virgin Islands. Atoll Research Bulletin 153:1-6. Roberts, C.M. 1993. Coral reefs: health, hazards and history. Trends in Ecology and Evolution 8:425-427. Roberts, C.M. 1995. Effects of fishing on the ecosystem structure of coral reef fishes. Conservation Biology 9:988-995. Roberts, C.M. 1997. Ecological advice for the global fisheries crisis. Trends in Ecology and Evolution 12:35-38. Roberts, C.M., and R.F.G. Ormond 1986. Habitat complexity and coral reef fish diversity and abundance on Red Sea fringing reefs. Marine Ecology Progress Series 41:1-8. Roberts, C.M., A.R.D. Shepherd, and R.F.G. Ormond 1992. Large-scale variation in assemblage structure of Red Sea butterflyfishes and angelfishes. Journal of Biogeography 19:239-250. Roberts, C.M., and N.V.C. Polunin 1993. Marine reserves: simple solutions to managing complex fisheries? Ambio 22:363-368. Roberts, C.M., and J.P. Hawkins 1997. How small can a marine reserve be and still be effective? Coral Reefs 16:150. Robins, C.R., and G.C. Ray 1986. A Field Guide to Atlantic Coast Fishes of North America. Peterson Field Guide Series: 32. Houghton Mifflin Co., Boston. 345 pp. Robertson, D.R., H.P.A. Sweatman, E.A. Fletcher, and M.G. Cleland 1976. Schooling as a mechanism for circumventing the territoriality of competitors. £co/ogy 57:1208-1220. Rogers, C.S., and J. Beets 2001 . Degradation of marine ecosystems and decline of fishery resources in marine protected areas in the US Virgin Islands. Environmental Conservation 28:312-322. Rogers, C.S., and V.H. Garrison 2001 . Ten years after the crime: lasting effects of damage from a cruise ship anchor on a coral reef in St. John, U.S. Virgin Islands. Bulletin of Marine Science 69:793-804. Rogers, C.S., L.N. McLain, and C.R. Tobias 1991. Effects of Hurricane Hugo (1989) on a coral reef in St. John, US VI. Marine Ecology Progress Series 78:189-199. Sladek Nowlis, J., and C.M. Roberts 1998. Fisheries benefits and optimal design of marine reserves. Fishery Bulletin 97:604-616. 584 Stokes, F.J. 1 984. Divers and Snorkelers Guide to the Fishes and Sea life of the Caribbean, Florida, Bahamas and Bermuda. The Academy of Natural Sciences, Philadelphia. 160 pp. U.S. Department of Commerce 2000. Census of population and housing: U.S. Virgin Islands. U.S. Census Bureau, Atlanta, Georgia. Population and Housing Profile 2000, 214 pp. Vine, P.J. 1974. Effects of algal grazing and aggressive behavior of the fishes Pomacentrus lividus and Acanthurus sohal on coral reef ecology. Marine Biology 24:131-136. Williams, A.H. 1979. Interference behavior and ecology of threespot damselfish {Eupomacentrus planifrons). Oecologia 38:223-230 585 Table 1. Site information for AGRRA fish surveys in the Virgin Islands. Site name Site Reef Latitude Longitude Survey Depth >25 cm % live stony # 30 m fish Total fish code Type'/ (0 . n N) (O'" w) Date(s) (m) stony corals coral cover transects species Exposure (^/lOm)" (mean ± sd)^ (# RDT)^ (#)“ Northern Virgin Islands St John (shallow) Great Lameshur, 12 F/W 18 18.853 64 43.312 May 26 00 3 8.5 18.5 ±3.5 10(1) 57 Donkey Great Lameshur, 13 F/W 18 19.094 64 43.390 Oct 23 99 5.5 4 13.5+6.0 10(1) 47 VIERS Fish Bay East, 11 F/W 18 19.073 64 45.808 Nov 14 99 5 3.5 10.516.0 10(1) 44 Inner Fish Bay West, 14 F/W 18 19.053 64 45.878 Aug 15 00, 5 8.5 7.016.0 10(1) 40 Inner Oct 24 00 St. John (deep) Great Lameshur, 15 F/W 18 18.572 64 43.302 Aug 3 99, 11 10.5 48.51 13.5 10(1) 70 Tektite Nov 29 99 Great Lameshur, 16 F/W 18 18.831 64 43.596 July 23 99 12.5 4.5 16.018.0 10(1) 59 Yawzi Fish Bay East, 9 F/W 18 18.948 64 45.777 Feb 08 00 6.5 4.5 12.015.0 10(1) 56 Outer Fish Bay West, 10 F/W 18 18.850 64 45.845 Aug 02 99 7.5 4 13.016.5 10(1) 47 Outer St. Thomas Brewer's Bay 2 F/W 18 20.670 64 59.157 May 26 99, 8.5 6.5 19.016.5 10(2) 58 (65)^ June 9 00 Buck Island 4 F/W 18 19.781 64 57.097 May 19 98, July 14 99 14 3.5 11.516.0 16(1) 71 Caret Bay 1 F/W 18 22.421 64 59.371 May 21 98, July 17 00 9.5 6 23.518.5 10(1) 58 Flat Cay 3 F/W 18 19.072 64 59.444 July 7 99, June 12 00 12 6 21.0 ±7.5 10(1) 51 Sprat Bay 5 FB/W 18 19.718 64 55.630 Aug 3 1 00 10.5 13 29.5 ±7.0 10(1) 54 Virgin Gorda Eustatia Reef 18 FB/W 18 30.50 64 20.250 July 24 00 9 9.5 17.016.0 10(1) 60 Guana Island Iguana Head 17 F/W 18 28.477 64 34.941 Aug 06 99 10 8.5 30.019.5 10(1) 64 St. Croix Cane Bay 6 F/W 17 46.230 64 48.530 Oct 14 99, Dec 16 99 10 7.5 30.51 11.0 12(3) 74 (92f Long Reef 8 FB/W 1746.118 64 41.490 Oct 13 99 13.5 7.5 16.013.5 24 (5) 68 (88)’ Salt River East 7 F/W 17 47.227 64 45.330 Oct 13 99 10 6 14.014.0 24 (6) 63 (117)’ Anegada Herman's Reef 19 BB/L 18 33.841 64 14.320 July 24 00 13 10 19.5 ±5.5 10(1) 53 Horseshoe Reef 20 B B/W 18 39.965 64 13.890 July 22 00 10.5 6 14.016.5 10(1) 49 Jack Bay 21 FB/W 18 44.961 64 19.246 July 22 00 9 4.5 8.014.5 10(1) 49 W. Cow Wreck 22 FB/W 18 45.164 64 24.596 July 23 00 8.5 7 11.5 ±3.5 10(1) 55 'F = fringing; F B = fringing barrier; B B = bank barrier; W = windward, L = leeward ^From Nemeth et al. (this volume) ^RDT = roving diver technique surveys “Number of fish species based on the maximum number seen on any one RDT survey. ^Number in parenthesis = cumulative number of species for multiple RDT surveys. 586 Table 2. Twenty-five most frequently sighted fish species during roving diver surveys for all sites combined in the Virgin Islands, with density (mean ± standard deviation) for species in belt transects. Fish Species Rank' Sighting Frequency^ Density (#/100m^) Thalassoma bifasciatum 1. 100 - Acanf burns coeruleus 2. 100 7.0 ±6.58 Sparisoma aurofrenatum 3. 100 4.4 ±3.03 Sparisoma viride 4. 100 3.6 ±2.42 Haetnulon flavolineatum 5. 100 3.3 ±3.90 Scarus taeriiopterus 6. 94.7 4.8 ±8.82 Stegastes partitus 7. 89.5 - Midloidichthys martinicus 8. 86.8 - Epinephelus cruentatus 9. 81.6 0.7 ± 1.04 Serrams tigrinus 10. 81.6 0 Acant burns babianus 11. 80.7 7.7 ± 7.94 Halicboeres garnoti 12. 80.7 - Cbaetodon capistratus 13. 80.7 1.0 ±0.95 Scarus croicensis 14. 78.1 3.6±3.12 Cbromis multilineata 15. 75.4 - Cbromis cyanea 16. 75.4 - Holocentrus rufus 17. 70.2 - Epinepbelus fulvus 18. 68.4 0.7 ± 1.03 Abudefduf saxatilis 19. 62.3 - Stegastes diencaeus 20. 59.6 - Cantbigaster rostrata 21. 58.8 - Corypbopterus glaucofraenum 22. 57.9 - Myripristis Jacobis 23. 57.0 - Stegastes planifrom 24. 56.1 - Halicboeres macidipinna 25. 56.1 - 'Rank of species with the same sighting frequency was determined by calculating the REEF density index for each species based upon the weighted abundance categories: S (single=l), F (few = 2), M (many = 3) and A (abundant = 4). Equation to calculate weighted density average was: Density = [(S*l) + (F*2) + (M*3) + (A*4)]/ (number of surveys in which species was observed) ^Sighting frequency (%SF) for each species was calculated using the equation: %SF = [S + F + M + A]/ (number of surveys) 587 Table 3. The total number and density (mean ± standard deviation) of the twenty-five most common species counted in belt transects for all sites combined in the Virgin Islands. Fish species Rank AGRRA fishes without <5 cm scarids + haemulids AGRRA fishes with <5 cm scarids + haemulids' Sum m Density (#/100 mb Sum (#) Density (#/100 mb Acanthurus bahianus 1 1176 7.7 ± 7.94 Acanthurus coeruleus 2 1070 7.0 ±6.58 Scarus taeniopterus 3 743 4.8 ±8.82 1640 10.5 Spahsoma aurofrenatum 4 677 4.4 ±3.03 911 6 Scarus croicensis 5 556 3.6 ±3.12 869 5.5 Sparisoma viride 6 549 3.6 ±2.42 920 6 Haetnulon flavolineatum 7 510 3.3 ±3.90 4926 32 Haemulon aurolineatum 8 309 2.0 ±5.35 311 2 Acanthurus chirurgus 9 294 1.9 ±5.05 Ocyurus chrysurus 10 204 1.3 ±2.17 Microspathodon chrysurus 11 191 1.2 ±2.11 Sparisoma rubripinne 12 153 1.0 ± 1.82 166 1 Chaetodon capistratus 13 147 1.0 ±0.95 Epinephelus cruentatus 14 113 0.7 ± 1.04 Epinephelus fulvus 15 113 0.7 ± 1.03 Haemulon plumieri 16 104 0.7 ± 1.75 108 0.7 Haemulon sciurus 17 100 0.6 ± 1.94 100 0.6 Lutjanus mahogani 18 93 0.6 ± 1.56 Scarus vetula 19 79 0.5 ±0.76 285 2.05 Caranx ruber 20 77 0.5 ± 1.06 Holocanthus ciliarus 21 74 0.5 ±0.95 Lutjanus apodus 22 74 0.5 ± 1.03 Lutjanus griseus 23 59 0.4 ± 1.40 Lutjanus jocu 24 57 0.4 ± 1.91 Melichthys niger 25 57 0.4 ± 1.12 Except St. Croix, where <5 cm scarids and haemulids were not counted. 588 Table 4. Density (mean ± standard deviation) of AGRRA fishes by site in the Virgin Islands. Site name Site code Herbivores Density (#/100 m^) Carnivores Macroalgal index Rel. (Abs.)^ Acanthuridae Scaridae (>5 cm) M. chrysurus ' Haemulidae (>5 cm) Lutjanidae Serranidae" Northern Virgin Islands St. John (shallow) Great Lameshu 12 7.5 ±7.0 10.8 ±7.3 0 28.5 + 25.8 2.5 ±5.5 0.6 ± 1.0 71 (80) Donkey Great Lameshur 13 8.0 ±5.0 17.7± 12.7 0 32.5 ±29.4 3.0 ±2.5 0.5 ±2.0 46 (72) VIERS Fish Bay East 11 28.5 ±50.5 14.0 ± 16.3 0.3 ± 1.0 32.3 ±23.1 10.5 ± 15.0 0.2 ±0.5 108(116) Inner Fish Bay West 14 8.5 ± 13.0 3.8 ±4.0 0 3.5 ±5.5 4.5 ±7.5 0.4 ± 1.0 278 (181) Inner St. John (deep) Great Lameshur 15 5.5 ±3.0 13.7 ± 10.8 0.2 ±0.5 5.0 ±5.5 10.0 ±29.5 2.5 ±2.0 190(112) Tektite Great Lameshur 16 10.0 ±4.5 48.8 ±30.1 0.2 ±0.5 15 ± 17.2 2.0 ±2.0 L0± 1.5 55 (31) Yawzi Fish Bay East 9 60.5 ±78.0 6.8 ±5.0 0 8.5 ± 17.3 3.0 ±3.5 0.4 ± 1.0 6(7) Outer Fish Bay West 10 45.0 ±57.0 22.2 ± 17.3 2.3 ± 1.4 7.0 ± 18.8 3.6 + 3.0 0.7 ± 1.0 52 (30) Outer St. Thomas Brewer's Bay 2 L0± 1.5 40.7 ±58.2 0 2.0 ± 4.3 1.0 ±2.0 0.5 ±0.5 83 (118) Buck Island 4 3.5 ±3.0 9.7 ±8.9 0.1 ±0.4 7.0 ± 17.4 3.0 ±5.5 2.0 ±2.5 173 (96) Caret Bay 1 9.5 ±5.0 14.5 ± 10.7 1.7 ± 1.4 0.5 ±0.7 0.5 ± 1.0 1.0 ± 1.0 110(82) Flat Cay 5 8.0 ±4.5 28.0 ± 13.9 0 2.7 ±3.5 2.0 ±3.0 1.5 ± 1.0 93 (53) Sprat Bay 5 4.0 ±3.0 17.8 ± 15.4 0.5 ±0.8 1.0± 1.4 1L5±31.0 1.0 ± 1.0 247 (201) Virgin Gorda Eustacia Reef 18 2L0±44.5 12.0 ± 12.8 4.3 ±3.1 1.8 ±2.9 5.5 ± 12.0 0.5 ±0.5 116(86) Guana Island Iguana Head 17 13.0 ± 15.0 13.0 ±5.3 0.7 ± 1.2 1.5 ±2.5 7.0 ±7.5 2.5 ±2.0 8(8) St. CroLx Cane Bay 6 5.5 ± 4.0 14.5 ±7.5 2.4 ± 1.9 9.5 ± 15.0 0.5 ± 1.0 5.5 ±2.5 35 (70) Long Reef 8 14.0 ±6.5 19.5 ± 13.0 1.6 ± 1.4 3.5 ±3.5 L5± 1.5 3.5 ±2.5 25 (19) Salt River East 7 14.5 ±8.5 10.0 ±6.0 0.4 ± 1.0 2.0 ± 1.5 0.5+ 1.0 3.0 ±2.5 3(2) Anegada Herman's Reef 19 13.5 ±6.0 27.7 ± 10.5 3.5 ±2.1 3.2 ±2.7 0 2.0 ±2.5 104 (55) Horseshoe Reef 20 40.5 ± 14.5 32.8 ±32.4 1.2 ±2.7 0 7.0±2L0 0.5 ± 1.0 147 (70) Jack Bay 21 26.0 ±23.5 10.8 ±7.6 9.0 ±6.7 4.5 ±5.2 7.5 ± 13.5 4.0 ±3.0 101 (77) W. Cow Wreck 22 32.5 ±9.5 6.8 ±4.5 0 5.5 ± 10.5 L5±3.5 1.5 ±2.5 230 (150) 'M chrysurus = Microspathodon chrysurus ^Epinephelus spp. and Mycteroperca spp. ^Rel = % relative macroalgal abundance x macroalgal height; Abs. = % absolute macroalgal abundance x macroalgal height; from Nemeth et al. (this volume). 589 Table 5. Length (mean ± standard deviation) in cm of AGRRA fishes by site in the Virgin Islands. Site name Site Acanthuridae Scaridae Haemulidae Lutjanidae Serranidae' Pomacanthidae Rank' code (> 5 cm) (> 5 cm) Northern Virgin Islands St. John (shallow) Great Lameshur 12 10.5 ±3.5 19.7 ± 13.5 16.1 ± 1.5 22.5 ± 4.5 24.5 ± 18.0 + + Donkey Great Lameshur 13 II.0±4.0 16.0 ± 1.0 I0.0±2.0 14.0 ±7.5 i3.0±0 + VIERS Fish Bay East 1 1 4.5 ±2.5 1 7.9 ±8.0 1 1.5 ±6.0 16.5 ±8.0 35.5 - -t- Inner Fish Bay West 14 9.5 ±6.0 1 8.0 ±4.0 16.9 ±4.0 20.5 ±9.5 19.0 ±23.5 25.5 ± I.O + Inner St. John ( deep) Great Lameshur 15 II.0±3.5 1 8.0 ±4.0 14.0 ±4.0 18.0 ± 14.0 17.0 ±6.0 9.0 ±9.0 .... Tektite Great Lameshur 16 11.4 ±2.5 I7.5± 1.5 14.5 ±6.0 20.0 ±6.5 23.5 ±7.5 13.5 ±3.0 Yawzi Fish Bay East 9 1 1.5 ±3.5 21 ±3.0 17.5 ±2.5 2L5±3.5 30.5 ±7.0 15.5 ±0 +++++ Outer Fish Bay West 10 1 0.0 ±2.2 20.5 ±5.5 17.4 ±3.0 20.0 ± 7.0 24.0 ± 10.5 9.0 ±9.0 - Outer St. Thomas Brewer's Bay 2 4.5 ±3.0 I0.0±4.0 I5.5±0 1 1.0 ±4.0 15.5 ±0.0 ... Buck Island 4 8.5 ±4.5 17.0 ±3.0 1 1.5 ±6.5 27.0 ± 18.5 I5± II.O 12.0 ±5.5 - ++ Caret Bay I 13.0 ±2.0 16.5 ±2.0 20.5 ±7.0 I5.5±0 19.0 ± 8.0 +++ Flat Cay 3 1 1.0 ±3.0 12.0 ±2.5 16.0 ±5.0 I5.5±0 17.0 ±4.0 13.5 ±2.5 Sprat Bay 5 I0.0±2.5 18.0 ± II.O 14.0 ±3.5 22.0 ±5.0 18.0 ± 10.5 25.5 Virgin Gorda Eustacia Reef 18 16.5 ±2.5 19.0 ±5.0 18.0 ±3.0 25.0 ±2.5 20.5 ±7.0 25.5 ±0 Guana Island Iguana Head 17 14.0 + 4.0 19.0 ±3.0 1 5.0 ±9.0 23.0 ±6.0 19.5 ±4.5 35.5 St. Croix Cane Bay 6 17.5 ±3.5 2L0±7.5 I7.5±2.5 15.5 ±0 19.0 ±4.0 + Long Reef 8 15.5 ± 1.5 19.0 ±3.5 1 8.5 ±3.0 22.5 ±5.0 20.0 ±3.5 1 8.0 ±6.0 -. + Salt River East 7 16.0 ± 1.5 22.0 ± 4.5 21.0 ±5.0 20.5 ±7.0 2I.0±4.0 I2.0±5.5 +++ Anegada Herman's Reef 19 13.7 ±2.0 16.8 + 2.0 22.0 ±5.0 26.0 ±2.5 14.0 ± 16.0 - Horseshoe Reef 20 10.5 ±2.0 1 7.3 ±2.0 33.5 ± 14.0 30.5 ± 18.0 15.5 ++ Jack Bay 21 14. 1 ±2.5 20.0 ±7.0 1 9.0 ±4.0 25.5 ±0 15.0 ±5.5 2.5 + W. Cow Wreck 22 8.0 ±3.0 2L0±5.0 23.0 ±4.5 25.5 ±0 20.5 ±4.5 7.0 ±6.0 — ++ ' Epinephelus spp. and Mycteroperca spp. ^Ranking = number of times a site within an island (i.e. St. Thomas) or group (e.g., St. John shallow) was represented by the smallest (-) or largest (+) value within the size range of a family. Angelfish were excluded due to the number of missing cells and Virgin Gorda and Guana Island were excluded since they only had one site per island. 590 lA Klein ^ Bonaire wKb Margin cO'IOOO m) Htothic tiu tixatlont Kllom«t«r$ Figure IB. Bonaire, Netherlands Antilles Figure 1. AGRRA surey sites at (A) Lighthouse Atoll, Belize and (B) off Bonaire ASSESSMENT TABLES FOR ABACO, BAHAMAS (FISH), LIGHTHOUSE ATOLL, BELIZE (CORALS, ALGAE, FISHES), AND BONAIRE, NETHERLANDS ANTILLES (CORALS, ALGAE, FISHES) BY PHILIP A. KRAMER,* AND BAERBEL G. BISCHOF INTRODUCTION Rapid assessments using the AGRRA protocol were conducted at three locations within the wider Caribbean during 1999. The results are presented here only as a series of data tables similar in format to other papers in this volume and were only used in the regional synthesis (Kramer, this volume). All data tables were generated from the AGRRA Access database using queries to generate site means based on either transect values (fish density, coral cover, coral density, Diadema density) or individual coral values (coral mortality, diameter, condition), or quadrat values (relative abundance of algae and canopy height of macroalgae). The location of the sites for Lighthouse Atoll and Bonaire are shown in figures 1 A and IB (see Figure 1 in Feingold et al., this volume, for the Abaco site locations). Tables 1-2 show results of the fish assessments collected in Abaco, Bahamas (for benthic results, see Feingold et al., this volume). Tables 3-6 show stony corals, algae, and fish results from Lighthouse Atoll, Belize, Tables 7-10 show stony corals, algae, and fish results from Bonaire, Netherlands Antilles. ' Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Sciences, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: pkramer@rsmas.miami.edu Table 1. Site information for AGRRA fish surveys in Abaco, Bahamas. 592 Table 2. Density (mean ± standard deviation) of AGRRA fishes and macroalgal index, by site for Abaco, Bahamas. Herbivores (#/100m^) Carnivores (#/100mh Macroalgal Acanthuridae Scaridae Haemulidae Lutjanidae Serranidae index 593 m in c/3 00 CJ u > d 00 Jq £ lU Ui O CO CQ T :S 3 O 0^ ’O ’O is x: t: o :z 1q X o CQ U o d o o X 3 O OO 3 1 3 O CO Q, > o lU CQ 3: cQ U CQ U T CQ SP u u u u o cQ oi -o "O u cQ CQ u CO o o cQ cQ u iC *0 O X) O X> O t O § § C >> C >> O CQ S 00 oo UU w u u. C/D s s u Table 3. Site information for AGRRA stony coral and algal surveys in Lighthouse Atoll, Belize. 594 o < (U c/2 3 o x: -C DO .S d) s .2 S o CN Al cd ;-i o o c o cd <+4 O _o cd ■> d) "O cd Td c cd -H C cd 33 (DX) .s a X) > GO 4H 3 3 e d U J3 CO C aC O < a 'c C3 h 3 cd E cd n: NO 00 > a- ’rt Jii < m 3 U Q u o o Long Caye Ridge 10 27.5 ± 7.5 27 2.0 ±1.9 9.8 ±4.5 2.8 ±2.5 3.0 ±2.6 3.8 ± 2.9 Marie’s Promenade 10 32.5 ± 10.5 23 9.0 ±3.4 4.3 ± 4.3 1.8 ±1.2 1.3 ±3.2 0.5 ± 0.8 Painted Wall 10 43.5 ± 14.0 25 8.8 ±14.6 14.7 ±4.8 3.5 ±1.7 5.0 ±6.9 4.3 ± 2.6 Silver Caves 10 25.0 ± 5.5 24 3.7 ±2.5 8.2 ±4.2 3.3 ±1.4 1.5 ±2.5 2.2 ±1.4 Surge City 10 26.5 ± 14.0 22 10.3 ±7.3 4.3 ±3.5 1.7±1.8 0.8±1.2 1.0±1.2 Table 7. Site information for AGRRA stony coral and algal surveys in Bonaire. 596 Table 9. Algal characteristics, and density of stony coral recruits and of Diadema antiilarum (mean ± standard deviation), by site in Bonaire. 597 598 Figure 1. AGRRA survey sites (boldface) in central-southern Quintana Roo, Mexico. CONDITION OF CORAL REEF ECOSYSTEMS IN CENTRAL-SOUTHERN QUINTANA ROO (PART 3: JUVENILE REEF FISHES) BY CARLOS GONZALEZ-SALAS,*’^ ENRIQUEZ NUNEZ-LARA,' MIGUEL A. RUIZ-ZARATE,‘ ROBERTO C. HERNANDEZ-LANDA,' and J. ERNESTO ARIAS-GONZALEZ' ABSTRACT The spatial patterns of coral reef fish recruits were assessed using a visual census method at three scales (subreefs, reefs, and areas) between June and August, 1999 in the spur-and-groove habitat of six fore reefs in the Mexican Caribbean. Six thousand three hundred twenty-seven fish recruits belonging to 54 species in 30 genera and 18 families were counted. Slight differences were found in composition and density of all species at all three spatial scales. A multiple regression analysis indicated statistically significant relationships with recruit density that were positive for mean diameter and negative for the live/dead ratio of “large” (>25 cm diameter) stony corals. Recruit density may depend largely on the intrinsic behavior of each species in direct relation to food and refuge availability, rather than on live coral coverage. INTRODUCTION The fringing coral reefs of the Mexican Caribbean form the northernmost part of the Mesoamerican Barrier Reef They extend 350 km along the eastern coast of Quintana Roo state from the northeasterly corner of the Yucatan Peninsula to its southern frontier with Belize (Fig. 1). The Sian Ka’an Biosphere Reserve is located in the central-eastern portion of Quintana Roo from just south of Tulum (20° 06' N, the northern boundary of Sian Ka’an), to Punta Punticub (19° 05’ N). The reserve consists of 528,147 hectares with limited access and includes Ascension and Espiritu Santo Bays. About 120,000 hectares are coastal and marine environments encompassing the coral reef zone (Gutierrez- Carbonel et al., 1993). Six reefs were sampled as part of the present study, three within the Sian Ka’an Biosphere Reserve and three outside the Reserve in the southern portion of the Mexican Caribbean (Fig. 1). Two of the reserve’s reefs are remote from population centers (Boca ' Laboratorio de Ecologia de Ecosistemas Arrecifes Coralinos, Centro de Investigacion y de Estudios Avanzados IPN, Carretera Antigua a Progreso Km 6, AP 73, Merida, Yucatan, Mexico. Email; earias@mda.cinvestav.mx 2 Ecole Pratique des Hautes Etudes, Laboratoire d’Ichtyoecologie Tropicale et Mediterraneenne, ESA 8046 CNRS, Universite de Perpignan, 66860 Perpignan cedex France. E-mail: cgonzale@univ-perp.fr 600 Paila, Punta Yuyum), which has allowed their natural conditions to be largely conserved. The third, Punta Allen, is situated near the settlement of Rojo G6mez where the principal activity is a small-scale lobster fishery. Two of the reefs outside of the reserve (Mahahual, Xcalak) are located near human settlements where small-scale fishing is the predominant activity although most of the fleet’s capture is focused on the open sea or at Chinchorro Bank rather than in the coastal reef zone. The reef at Xahuayxol has been relatively undisturbed anthropogenically. The principal fishing methods used are fixed traps, trotlines, gill nets and harpoons and these are mostly directed toward capture of commercially important species such as grouper (serranids), barracuda (sphyranids), snapper (lutjanids) and grunts (haemulids). Relatively little research has been done on the coral reefs in the Mexican Caribbean and even less on its reef fishes. Earlier studies, predominantly focused on adult fish community structure, have been summarized by Arias-Gonzalez et al. (1997) and Salazar et al. (1997). Recently Castro (1998) has characterized the fish community structure of Mahahual Reef Nuhez-Lara (1998) and Nuhez-Lara and Arias-Gonzalez (1998) have demonstrated the importance of environmental factors, particularly topographic complexity, in determining reef fish community structure. Arias-Gonzalez (1998) has created trophic models for a protected and an unprotected zone, finding important differences reflected in fish biomass. Diaz-Ruiz et al. (1999) presented evidence that variations in species diversity and trophic structure are associated with sequential habitat use during the life cycle. Recruitment is widely considered a key structuring process in reef fish communities (Doherty and Williams, 1988; Doherty and Fowler, 1994). Previous studies in the Western Atlantic have focused on the factors that regulate temporal and spatial variability in recruitment such as predation, refuge availability, and habitat use (Shulman, 1985a, 1985b; Shultz and Cowen, 1994; Booth and Beretta, 1994; Caselle and Warner, 1996). Enhanced understanding of natural fluctuations in recruitment and its relationship to coral reef conditions would allow more effective management of fishery resources. This study is the first assessment of coral reef fish recruitment patterns at different spatial scales in the Mexican Caribbean. Relationships between recruitment and descriptors of benthic reef condition are also examined. METHODS Reef fish recruitment patterns were visually assessed between June 25 and August 31, 1999 on three spatial scales: area, reef and subreef The central and southern areas of Quintana Roo have coastlines that are approximately 100 km long. Each of these areas was represented by three, strategically chosen reefs (see Ruiz et al., this volume), which were separated from one another by about 25-30 km. Every reef was partitioned into three representative subreefs at approximately 1 km intervals. Ten belt transects, each 30 m long by 1 m wide, were swum parallel to the coast at a depth of 12 m in the spur-and- groove habitat of the fore-reef zone by one diver (Gonzalez-Salas) in every subreef The spacing between adjacent transects was about 50 m. To assess the diurnal community of juvenile reef fishes, all surveys were made between 09:00 and 14:00 hours. Recruits were identified to species, or to the lowest possible taxonomic level. Their size was estimated 601 with a T-bar. Fish identifications were based on the descriptions of Randall (1983), Lindeman (1986), Humman (1994), and Lieske and Myers (1995). Recruitment patterns were analyzed on the three spatial scales. Multiple regression analysis was used to estimate the relationship between juvenile density and benthic variables [total live stony coral cover; total, old, and recent partial-colony mortality, live/dead ratio, mean diameter, and percent bleached or diseased colonies for “large” (>25 cm diameter) stony corals; relative abundance of macroalgae and turf], which were measured in the same sites by Ruiz et al. (this volume). To evaluate the affinity of the sampling stations, a multivariate classification analysis based on the density of the fish species was made using the Bray-Curtis distance index (Bray and Curtis, 1957), complemented by the Unweighted Pair Grouping Method Average (UPGMA) cluster method. The Atlantic and Gulf Rapid Reef Assessment (AGRRA) fishes constitute a subset of the “all species” data: in this paper, “serranids” are species of Epinephelus (excluding E. cruentatus and E.fulvus, here considered to be Cephalopholis cruentata and C. fulva, respectively) and Mycteroperca. RESULTS A total of 54 species of reef fish recruits belonging to 30 different genera and 18 families were identified in the surveys. No difference was noted in species richness at the largest spatial scale (Fig. 2), but the central area had reefs with both the highest (Boca Paila) and lowest (Punta Allen) numbers of species (Table 1). A total of 6,327 reef fish recruits were counted: 3,337 in the central area and 2,990 in the southern area. Flence, the “all species” recruit density was somewhat higher in the central area (Fig. 3). The highest densities, found in Boca Paila (central area) and Xcalak (southern area), were about twice that of the lowest in Xahuayxol (southern area). The mean “all species” recruit density overall was 117 individuals/ lOOm^. The 25 most frequently sighted species (Table 2) were a mixture of herbivores (Stegastes, Span soma, Acanthurus, Scarus) and carnivores that feed primarily on the benthos (e.g., Halichoeres, many other genera). Overall, recruits of the AGRRA herbivores (scarids, acanthurids) were more abundant than the AGRRA carnivores (lutjanids>haemulids>serranids as defined above) (Table 3). Mean densities of parrotfish (scarid) recruits were highest in the two most northerly reefs of the central area (Boca Paila, Punta Yuyum) and in Mahahual, the most northerly reef of the southern area. Surgeonfish (acanthurid) recruits were most common in the most southerly reef (Xcalak) and least abundant in Mahahual. The multiple regression analysis indicated that the mean diameter and live/dead ratio of the >25 cm in diameter stony corals together explained more than 60% of the variability in recruit density. The relationship with mean coral diameter was positive (r^= 0.276, p = 0.02), whereas that with the live/dead ratio was negative (r^= 0.379, p< 0.05) (Fig. 4). The remaining variables had no significant relationship with recruit density. A numerical classification analysis in Q mode identified five groups of fish recruits (Fig. 5). The first includes four subreefs from the southern area and one from the central area. The second and fifth groups are restricted to the central area whereas the third and fourth are located in the southern area. 602 Mexican Caribbean Mean species richness/reef in central region 45 , Mean species richness/reef in southern region 32 i/i *3 0^ a irx © 39 31 28 29 36 31 35 i BPN BPC BPS PYN PYC PYS PAN PAC PAS MN MC MS XN XC XS XCN XCC XCS SUBREEFS Figure 2. Species richness of all species of juvenile coral reef fishes, by subreef in central-southern Quintana Roo, Mexico. See Table 1 for site codes. Mexican Caribbean 124 111 Mean density / reef Mean density / reef 300 140 126 104 125 71 135 I 250 o % 200 BPN BPC BPS PYN PYC PYS PAN PAC PAS MN MC MS XN XC XS XCN XCC XCS SUBREEFS Figure 3. Mean density (no. individuals/ 100 m^) of all species of juvenile coral reef fishess by subreef in central-southern Quintana Roo, Mexico. See Table 1 for site codes. 603 A Diameter C 'V ‘c s B Live:dead ratio Figure 4. Regression plot between mean juvenile reef fish density (no. individuals/ 100 m^) and (A) mean stony coral diameter (y=l. 041+0. 00795374x, r^=0.276, f(l,16)=6.13, p=.002) and (B) mean stony coral live/dead ratio (y=l. 73079-0. 15325x, r^=0.379, f(l,16)=9.79, p=.006), by subreef in central- southern Quintana Roo, Mexico. DISCUSSION The similarity in the species richness of recruits in the central and southern areas is probably attributable in part to the great similarity in their reef structures although large-scale physical processes (such as current patterns) should also be taken into account. Much of the subreef- and reef-scale variability in reef fish recruitment could be related to local, coastal environmental factors. For example, the relatively high values found off Boca Baila (northern area) and off Xahuayxol and Xcalak (southern area) were all in reefs that are close to lagoons or bays that may serve as sources of recruits. 604 Figure 5. Hierarchical classification analysis of the juvenile fish transect data by subreef in central- southern Quintana Roo, Mexico. See Table 1 for site codes. Labels A-E represent groups of sites with similar juvenile fish community structure. Whereas adult herbivores (parrotfishes and surgeonfishes) were relatively abundant in the more heavily fished reefs where their natural predators are less common (Nunez-Lara et al., this volume), the mean density of their recruits showed relatively little between-reef variation (Table 3). The relative paucity of juvenile snappers and grunts compared to adult densities found by Nunez-Lara et al. (this volume) is probably related to the strong association of the juveniles with lagoonal systems, particularly in Boca Paila, Punta Yuyum and Punta Allen. Recruitment variation at any of the three scales is likely to be influenced by one or more of the following factors; differential larval availability; differential settlement during habitat selection; mortality differences in early, post-settlement stages; or post- recruitment movement towards different preferred habitats (Caselle, 1996). The positive relationship between recruit density and stony coral diameter suggests that the survival of reef fish recruits is enhanced by the greater structural complexity of large corals. Indeed, Nemeth (1998) has recently documented the positive effects of hole and crevice density on recruit abundance. The overall negative relationship with the live/dead coral ratio, as previously found in St. Croix, U.S. Virgin Islands (Booth and Beretta, 1994; Caselle and Warner, 1996) is an indication that recruit density is largely independent of the condition of these stony corals. In other words, the dominant factors affecting reef fish recruitment apparently are not associated with the benthos condition indicators recorded by Ruiz et al. (this volume) but rather with larval input and behavior, as well as the availability of refuges and food. If the scales at which recruitment is being carried out are smaller than those employed in this study, it will be necessary to measure other variables, such as refuge type, or the heterogeneity, size and density of refuges in reef substrata. 605 ACKNOWLEDGMENTS This project was financed by the Caribbean Environment Programme of the United Nations Environment Programme, Consejo Nacional de Ciencia y Tecnologia (CONACYT) and Centro de Investigacion y de Estudios Avanzados (CINVESTAV). The AGRRA Organizing Committee also facilitated financial support as described in the Forward to this volume. We express our sincere thanks to the Secretaria de Medio Ambiente, Recursos Naturales y Pesca (SEMARNAP) in Quintana Roo for allowing us to use their facilities at Mahahual. Thanks are also due to Biol. Alfredo Arrellano Guillermo, Director, Sian Ka’an Biosphere Reserve (RBSK) for his invaluable logistical assistance during the work within the Reserve, to Amigos de Sian Ka’an for permission to use Pez Maya as a work base within the Reserve, and to Chente, Chepe, Felipe, Fredy and Vidal for their valuable help during the fieldwork. REFERENCES Arias-Gonzalez E., E. Nunez-Lara, C. Gonzalez-Salas, and R. Salazar-Murguia 1997. Trophic structure of fish communities: a comparison of reef ecosystems of Gulf of Mexico and the Mexican Caribbean. Simposio Ecosistemas Acuaticos de Mexico, 23 pp. Arias-Gonzalez, J.E. 1998. Trophic models of protected and unprotected coral reef ecosystems in the south of the Mexican Caribbean. Journal of Fish Biology. 53 (supplement A): 236-255. Booth, D.J., and G.A. Beretta 1994. Seasonal recruitment, habitat associations and survival of pomacentrid reef fish in the US Virgin Island. Coral Reefs 13:81-89. Bray, J.R., and J.T. Curtis 1957. An ordination of the upland forest communities of southern Wisconsin. Ecological Monographs 27:325-349. Caselle, J.E., and R. Warner 1996. Variability in recruitment of coral reef fishes: the importance of habitat at two spatial scales. Ecology. 77:2488-2504. Castro, J.E. 1998. Estructura de la Comunidad de peces asociada al Arrecife Mahahual, Quintana Roo, Mexico. Tesis de Maestria. CINVESTAV I.P.N Unidad B Merida, Yucatan. 96 pp. Diaz-Ruiz, S., A. Aguirre-Le6n, and J.E. Arias-Gonzalez 1998. Habitat interdependence in coral reef systems: a case of study in a Mexican Caribbean Reefs. Journal of Aquatic Health and Management Vol: 1 :387-400. 606 Doherty P., and D. McB, Williams 1988. The replenishment of coral reef fish populations. Oceanography and Marine Biology Annual Review 26:487-55 1 . Doherty P, and A.J. Fowler 1 994. An empirical test of recruitment limitation in a coral reef fish. Science 263:935-939. Eschmeyer,W.N. (editor) 1998. Catalog of Fishes. Special Publication No. 1 of the Center for Biodiversity Research and Information, California Academy of Science, San Francisco, CA. Volumes 1-3, 2905 pp. Gutierrez-Carbonel, D., S. Garcia, M. Lara, and C. Padilla 1993. Caracterizacion de los arrecifes coralinos de la Reserva de la Biosfera de Sian Ka’an, Quintana Roo. SIAN KA ’AN series. Documentos. Num. 1. 8 pp. Humman, P. 1 994. Reef Fish Identification: Florida, Caribbean, Bahamas. 2nd. Edition. Jacksonville, Florida. New World Publication, Inc. 396 pp. Liskie, E., and R.F. Myers 1995. Guide des Poissons des Recifes Coraliens: Region Caraibe, Ocean Indien, Ocean Pacifique, Mer Rouge. Paris, Delachaux et Niestle. 400 pp. Lindeman K.C. 1 986. Development of larvae of french grunt Haemulon flavolineatum, and comparative development of twelve western Atlantic species of Haemulon (Percoidei, Haemulidae). Bulletin of Marine Sciience 39:673-716. Nemeth, R. 1998. The effect of natural variation in substrate architecture on the survival of juvenile bicolor damselfish. Environmental Biology of Fishes 53:129-141. Nunez-Lara, E. 1998. Factores que Determinan la Estructura de la Comunidad de Peces Arrecifales en el Sur del Caribe Mexicano: un Analisis Multivariado. Tesis de Maestria. CINVESTAV I.P.N. Merida, Yucatan. 1 15 pp. Nunez-Lara, E., and E. Arias-Gonzalez 1998. Relationship between reef fish community structure and enviromental variables in the southern Mexican Caribbean. Journal of Fish Biology 53:209-221. Randall, J.E. 1983. Caribbean Reef Fishes. 2"^ edition. T.F.H. Publications, Inc. U.S.A. 350 pp. Salazar-Murguia R., C. Gonzalez-Salas, and J.E. Arias-Gonzalez 1997. Efectos de un area semi-protegida y no protegida del sur del Caribe Mexicano, sobre la estructura comunitaria de peces arrecifales. 50^^ Gulf and Caribbean Fisheries Institute 50:354-371. Shulman, M.J. 1985a. Variability in recruitment of coral reef fishes. Journal of Experimental Marine Ecology and Biology 89:205-219. 607 Shulman, MJ. 1985b. Recruitment of coral reef fishes.’ effects of distribution of predators and shelter. Ecology 66:1056-1066. Shultz, E.T., and R.K. Cowen 1994. Recruitment of coral reef fishes to Bermuda: local retention or long-distance transport? Marine Ecology Progress Series: 1 09: 1 5-28. Table 1. Site information for AGRRA juvenile fish surveys in central-southern Quintana Roo, Mexico 608 609 Table 2. Sighting frequency and mean density of the 25 most frequently sighted juvenile fish species in the “all species” belt transect surveys in central-southern Quintana Roo, Mexico, * = AGRRA species. Species name Sighting frequency (%)' Density (#/100m^) Halichores garnoti 83 13.20 Stegastes partitus 76 13.37 *Spansoma aurofrenatum 64 5.28 Thalasoma bifasciatum 57 13.54 Chromis cyanea 52 26.07 *Acanthurus coerules 51 3.04 *Acanthurus bahianus 48 3.31 *Sparisoma viride 35 1.69 *Cephalopholis fulva 32 1.44 (=Epmephelus fulvus) ^ *Holacanthus tricolor 32 1.33 Cantigaster rostrata 29 1.30 Stegastes dorsopunicans 28 1.50 *Scarus iserti (=S. croicensisf 28 3.09 Stegastes planifrons 28 1.93 *Bodianus rufus 22 1.00 *Cephalopholis cruentata 22 0.96 (=Epinephelus cruentatus) ^ Clepticus parrae 19 17.50 Halichoeres maculipinna 14 0.57 *Chaetodon capistratus 12 0.63 Stegastes variabilis 11 0.43 Pseudopeneus maculatus 11 0.39 Gramma loreto 11 0.72 *Sparisoma radians 10 0.76 *Sparisoma atomarium 10 0.48 *Lutjanus apodos 9 1.00 'sighting frequency (%) = percentage of transects in which the species was recorded. ^Genus and/or species names according to Eschmeyer’s (1998) revision. 610 Table 3. Density (mean ± standard deviation) of juvenile AGRRA fishes by site in central- southern Quintana Roo, Mexico. Site name Herbivores (#/100mh Carnivores (#/100mh Macroalgal Acanthuridae Scaridae Haemulidae Lutjanidae Serranidae' Index^ Central Boca Paila N 2.7 ±3.2 4.1 ±5.0 0.2 ±4.0 90 Boca Paila C 3.6 ±3.0 3.5 ±3.0 130 Boca Paila S 1.8 ± 1.0 2.6 ±3.0 113 Punta Yuyum N 2.7 ±2.0 4.6 ±4.0 0.003 96 Punta Yuyum C 1.6 ±4.0 5.8 ±3.0 7.3 ±0.7 81 Punta Yuyum S 2.5 ±0.7 2.5 ±3.0 81 Punta Allen N 2.5±0.7 2.9 ±2.0 142 Punta Allen C 4.0 ± 2.0 3.5± l.O 0.003 116 Punta Allen S 3.0 ±0.9 2.7 ±2.0 126 Southern Mahahual N 0.3 1.4 ±0.7 0.007 64 Mahahual C 0.3 4.6 ±3.0 108 Mahahual S 1.3 5.9 ±8.0 74 Xahuayxol N 3.6 ±0.9 1.5 ± 1.0 0.003 0.003 75 Xahuayxol C 3.7 ±3.0 1.4 ± 1.0 0.003 0.003 83 Xahuayxol S 1.6 ±8.0 1.0 ±0.4 0.007 71 Xcalak N 4.7 ±3.0 1.9 ± 2.0 0.007 52 Xcalak C 4.6 ±0.9 3.0 ±3.0 49 Xcalak S 4.2 ±2.0 2.6 ±3.0 0.013 0.017 53 ' Epinephelus spp. (excluding any E. cruentatus and E. fulvus) and Mycteroperca spp. ^From Ruiz et al. (this volume) APPENDIX ONE THE ATLANTIC AND GULF RAPID REEF ASSESSMENT (AGRRA) PROTOCOLS: FORMER VERSION 2.2 BY PATRICIA RICHARDS KRAMER* and JUDITH C. LANG^ INTRODUCTION The AGRRA methodology is the result of an on-going international collaboration of reef scientists and managers. Since its initiation in 1995, the methods have undergone a series of iterations (Table 1) in response to recommendations from the organizing committee and our colleagues. Version 2.2 of the AGRRA method is briefly summarized below as it was the version used for many of the assessments reported in this volume. Sections that have been changed in subsequent versions are underlined. Specific deviations from this or other versions of the protocols are detailed in the individual papers of this volume. For more information on the current version of the methodology, data-sheet templates, survey equipment, and the AGRRA approach, see the AGRRA website (http://coral.aoml.noaa.gov/agra/method/methodhome.htm). One of the main objectives of the AGRRA approach is to provide a standardized methodology enabling teams working in different areas to collect and compare data on a regional scale. With visual censuses, it is particularly important to minimize individual bias among observers. Hence, it is essential to carefully standardize methods prior to data collection. Suggestions for consistency training and calibration can also be found at the AGRRA website. SELECTION OF REEFS AND SITES For the purposes of AGRRA, a region is defined as the coarsest scale category (-100-1000 km scale); followed by an area (~10-100 km scale); a reef (-1-10 km scale); and a site (0.2 km scale). The method for selection of reefs to assess will be influenced in part by their local abundance and distribution and by the sampling effort to be undertaken. Whenever the extent and/or number are too large for complete sampling, the reefs should be subdivided or “stratified” and examples selected from within each subdivision. All good sources of information (benthic maps, aerial photographs, remote images, charts, local knowledge, recormaissance by Manta tow-board, etc.) should be employed to separate the reefs into * Marine Geology and Geophysics, Rosenstiel School of Marine and Atmospheric Science, University of Miami, 4600 Rickenbacker Causeway, Miami, FL, 33149. Email: kramer_p(@bellsouth.net ^ P. O. Box 539, Ophelia, VA 22530. Email: JandL@rivnet.net 612 subdivisions on the basis of geographic distribution and by secondary characteristics such as size, depth range, and position relative to land. If there are no clear bases for making subdivisions (e.g., a continuous bank barrier or fringing reef several kilometers long), a numbered grid 200 m long can be superimposed along a chart or image of the reef, with each number representing a potential survey site. Hardgrounds, pavements and other habitats that lack a framework constructed of reef-building corals should be avoided. Once reefs are stratified, depending on the methods and resources available for use, the ones that are selected for survey will generally fall into one of three categories: 1 . Unbiased-chosen on the basis of a random sampling strategy; 2. Representative-chosen with local knowledge to be representative of reefs in the area; 3. Strategic-chosen with local knowledge and for a purpose (e.g., they are considered threatened, degraded or in particularly good condition). For regional comparisons, it is best to have reefs that are chosen by either category (1) or (2). Reefs chosen by (3) should be clearly flagged as such. A site is defined as an area of habitat of 200 m x 200 m (or less) that is more or less homogeneous and accessible from a boat anchored or moored in one place. Two habitats of maximum reef development of particular interest are the ~ 1-5 m depth interval (shallow Acropora palmata zone) and the ~8-15 m depth interval (shallow fore reef or equivalent). Whenever possible, one site within each of these depth intervals should be surveyed on each reef chosen for an AGRRA assessment. The site description for each survey site should include an explanation of how the site was selected. BENTHOS (STONY CORALS, ALGAE AND DIADEMA) 1 . At each site, haphazardly lay a 1 0-m transect line (marked in 2 m intervals) just above the reef surface (= taut) in a direction that is approximately parallel to the long axis of the reef. Be sure to avoid or cross any other transects. Stay away from the edges of the reef and try to avoid areas with abrupt changes in slope, deep grooves, large patches of sand or unconsolidated coral rubble. Unusual reef features should be included only to the extent appropriate to their relative abundance at the site. 2. Record total live stony coral (scleractinians + Millepora) cover under the transect line (to the nearest 10 cm) with a measuring tool (e.g., a 1-m long PVC pipe marked in 10 cm intervals). If the reefs are too small to avoid sandy patches, record how much of the line crosses sand (for later calculation of live stony coral cover/m of reef hard reef substratum). 3. Swim along the transect line, stopping at the first stony coral (Table 2) located directly beneath the line that is at least 25 cm (or. if preferred, at least 10 cm) long and in original growth position or, if fallen, has either become reattached to the substratum or is too large to move. Record each of the following: A. Name (genus and species) B. The water depth at the top of the colony at the beginning and end of each transect and at any major change in depth (>lm). 613 C. After identifying the colony boundaries based on common skeletal connections among the living polyps and similarities in their size and color, record maximum colony projected diameter (live + dead areas) perpendicular to the axis of growth and maximum height (live + dead areas) above the substratum parallel to the axis of growth. D. Percentage estimates of the coral that are “recently dead” (corallite structures are white and either still intact or covered by a layer of algae or fine mud; include fish bites) and “old dead” (corallite structures are either gone or covered by organisms that are not easily removed, including brown encrusting clionid sponges) as viewed from above at an angle that is parallel to the axis of growth. If the entire coral has been dead a long time and can still be identified, at least to generic level based on gross (e.g., Acropora palmatd) or skeletal (e.g., Diploria) morphology, score as 100% “old dead.” E. Evidence anywhere on the entire colony of- • Any diseases, using the following color categories: BB = Black band WB = White band WS = White spots, patches or pox WP = White plague YB = Yellow blotch (sometimes called yellow band) RB = Red band UK = Unknown Underline any disease visible on the upper colony surface that contributed to the estimate of “% recent dead.” • Bleached tissues as approximate severity of discoloration; P = Pale (discoloration of coral tissue) PB = Partly bleached (patches of fully bleached or white tissue) BE = Bleached (tissue is totally white, no zooxanthellae visible) • All other sources of recent mortality that can still be unambiguously identified, such as sediment, storm damage, parrotfish bites, damselfish bites and/or algal gardens, predation (e.g., by Corallophilia abbreviata or Hermodice carunculata), and spatial competitors (e.g., benthic algae, invertebrates like Erythropodium caribaeorum or other stony corals). Underline any that contributed to the estimate of “% recent dead.” • Algal gardens established by damselfish (in live parts of the coral as numbers of resident fishes and/or the presence of their gardens). For large clusters or thickets in which colony boundaries are not distinguishable, use a standard point-count method to identify recent death, old death, or living coral tissue every 25 cm along the line. The maximum diameter and height should be determined for the entire cluster or thicket. 4. Go to the next appropriately sized coral and repeat step 3 above. Continue evaluating each coral until reaching the other end of the transect line. 614 5. Reswim the transect with the 25 x 25 cm algal quadrat. Starting at 1 m, place the quadrat every two meters directly below the transect line (i.e., at 1,355,7,9 m). If any of these area are unsuitable (i.e., <80% is covered by algae of any functional groupX place the quadrat on the nearest available space within a 1 m radius of the mark. If no suitable areas are present, draw a line through the space on the data sheet. For each quadrat that is in a suitable area, remove any thin layers of sediment that could cover crustose coralline algae and record each of the following: A. Subtratum type; B. % of macroalgae (all larger erect fleshy algae >1 cm in height: both fleshy and calcareous); C. % of algal turfs (mostly tiny filaments. <1 cm in height), including any below the canopies of macroalgae: D. % living crustose coralline algae (solid, calcareous encrusters that are pink or reddish in color, include any that are clearly visible below turf algae or a thin layer of sediment): E. By using the plastic ruler for scale, the average canopy height of the macroalgae in the quadrat. F. Optional-the number of all small (up to 2 cm maximum diameter) stony corals (scleractinians and Millepora) in the quadrat and, whenever possible, their identity to the genus level. 6. Swim along the transect, counting every Diadema antillarum (juveniles and adults) that can be seen within a belt extending 1/2 m on each side of the line. 7. Collect the line and haphazardly reset it at least 1 m laterally away from its previous position, following previous guidelines. Try to distribute transects around the survey site. 8. Repeat steps 2-6 for each transect. Continue to reset transects until a minimum of 100 corals and 50 quadrats have been assessed. 9. Enter transect data into the AGRRA benthos spreadsheet. After checking for accuracy, submit an electronic copy to agrra@rsmas.miami.edu. Back up data regularly and store in a safe place. FISH The following two distinct assessment methods should be applied at each site. Whenever possible, they should be conducted between 1000 and 1400 hours. As many fishes are wary of humans, it is important to try to keep away from other divers. 615 Method I. Belt Transects for AGRRA Species. All belt transects used for fish assessments should be located within the same general habitats and depth intervals (i.e., at 1-5 m and 8-15 m) as the benthic transects, but they will tend to be further apart and may range into deeper and shallower water. 1 . Lay a 30 m transect by first placing the weighted end of the tape on the bottom and then swimming in a straight line (by periodically fixing on a distant object) while releasing the tape from the reel, which can be clipped to the weight belt for easy release. 2. Swim at a more or less constant rate holding a 1-m wide T-bar (with 5 cm increments for scale) angled downward at about 45 degrees while looking consistently about 2 m ahead of the T-bar. Within a belt visually estimated to be 2 m wide, count only the fish listed in Table 3 (also listed on the data sheet), while using the T-bar to estimate the size of each fish within one of the following size categories (<5, 5-10, 10-20, 20- 30, 30-40, >40cm). If necessary, pause while recording data and then resume swimming. Do not count juvenile parrotfishes and grunts <5 cm in total length. Large, intraspecific groups can be classified into one-or-more size categories as necessary. By keeping an equivalent effort on all segments of the transect, the tendency to count all members of a school crossing the transect, instead of just those members which happen to be within the belt as counting of its segment takes place, can be avoided. 3. Recoil the tape and haphazardly redeploy at a distance of least 5 m laterally away from the previous position. 4. Repeat steps 1-3 for each transect. Continue to reset transects until a minimum of 10 transects have been completed. 5. Enter belt-transect data into the AGRRA fish spreadsheet. After checking for accuracy, submit an electronic copy to agrra@rsmas.miami.edu. Back up data regularly and store in a safe place. Modifications. Researchers wanting to census other species of fish are encouraged to do so on a separate pass over the transect, after the completing counts of AGRRA fishes. Method II. Roving Diver Technique After finishing the belt transects (or concurrently depending on the number of surveyors), a census of all species of fishes should be conducted in the same general depth and habitat as the belt transects, following the methodology of the Reef Environmental Education Foundation (REEF) (see http;//www.reef.org/) and briefly explained below. 616 1. Swim around the site for approximately 30 minutes (preferably 45-60 minutes), searching under overhangs, in caves and so on to find as many fish species as possible. 2. By the end of the dive, estimate the density of each species by using logarithmic categories: Single (1 fish). Few (2-10 fishes). Many (11-100 fishes), or Abundant (>100 fishes) and summarize these observations on a standardized REEF data entry sheet. 3. Transcribe the data on a standardized REEF Scantron sheet and send to REEF FIQ, at P.O. Box 246, Key Largo, FL 33037, USA. OPTIONAL COMPONENTS Stationary Plot Fish Survey The Stationary Plot Technique (Bohnsack and Bannerot, 1986), outlined below, has been used extensively in the Caribbean and the Florida Keys to provide data on abundance and size for a wide range of fish species. Its use is encouraged as a third way to quantify fishes at each site, but not as a replacement for Method I or II. 1. Count the number of fish observed in a visually estimated cylinder (radius of 7.5 m) for a period of five 5 minutes. 2. Estimate and record the length of each fish counted. Herbivory The Fish Bite Method (Steneck, 1985) can be used to gauge the effect of certain herbivorous fishes on algal composition in the survey sites. Herbivorous fish guilds are categorized as: Scrapers = Scaridae (parrotfish) Browsers = Acanthuridae (surgeonfish), Microspathodon chrysurus (yellowtail damselfish) Non-denuders = other Pomacentridae (damselfish) 1 . Use a metric scale in conjunction with natural landmarks on the reef surface (e.g., a small coral or gorgonian) to haphazardly delineate an area that is approximately 1 m square and representative of the benthic cover at the site. (Do not place a meter quadrat on the substratum as some fish are particularly prone to biting novel objects placed within their feeding territories). Back off as far as possible while still being able to see the meter-square area. Watch for five minutes. Record the depth, time of day, and number of bites from all species of fishes in the three guilds listed above, whenever possible identifying them to 617 species. It is necessary to be able to distinguish (a) juvenile scarids from other fishes with similar stripes, such as acanthurids and labrids (wrasses which only look as though they are biting algae as they search for amphipods to eat) and (b) yellowtail damselfish (which are browsers) from the species of damselfish that cultivate algal gardens. 2. Repeat for a total of five quadrats (and ~25 minutes of observation). ACKNOWLEDGMENTS Ken Marks provided the species names for Tables 1 and 2. We thank all of the AGRRA contributors for their suggestions on improving this version of the method. P.R. Kramer thanks E. Fisher, B. Ripkey, A.R. Lees, P.K. Kramer and M.K. Richards for technical support. J. Lang thanks G. Arcilo, D. Brewer, and the staff of the Akumal Dive Shop for field support in Akumal, Mexico during early field tests of the AGRRA protocol. REFERENCES Bohnsack, J.A., and S.P. Bannerol 1986. A stationary visual census technique for quantitatively assessing community structure of coral reef fishes. NOAA Technical Report National Fish and Wildlife Service 41:1-15. Cairns, S.D., D.R. Calder, A. Brinkmann-Voss, C.B. Castro, P.R. Pugh, C.E. Cutress, W.C. Jaap, D.G. Fautin, R.J. Larson, G. Richard Harbison, M.N. Arai, and D.M. Opresko 1991. Common and Scientific Names of Aquatic Invertebrates from the United States and Canada. Cnidaria and Ctenophora. American Fisheries Society Special Publication 22. Bethesda, Maryland. 75 pp. Eschmeyer,W.N. (editor) 1998. Catalog of Fishes. Special Publication No. 1 of the Center for Biodiversity Research and Information, California Academy of Science, San Francisco, CA. Volumes 1-3, 2905 pp. Foster, A.B. 1987. Neogene paleontology in the northern Dominican Republic. 4. The Genus Stephanocoenia (Anthozoa: Scleractinia: Astrocoeniidae). Bulletins of American Paleontology 93:5-22. Ginsburg, R.N., R.P.M. Bak, W.E. Kiene, E. Gischler, and V. Kosmynin. 1996. Rapid assessment of reef condition using coral vitality. Reef Encounter 19:12-14. Robbins, C.R. and R. M. Bailey, C. E. Bond, J. R. Brooker, E. A. Lachner, R. N. Lea, and W. B. Scott 1991. Common and Scientific Names of Fishes from the United States and Canada. Fifth Edition. American Fisheries Society American Fisheries Society Special Publication 20. Bethesda, Maryland. 183 pp. 618 Steneck, R.S. 1985. Adaptations of crustose coralline algae to herbivory: patterns in space and time. Pp. 353-356. In: D.F. Toomey and M.H. Nitecki (eds.), Paleoalgology: Contemporary Research and Applications . Springer- Verlag, Heidelberg. Weil, E., and N. Knowlton 1994. A multi-character analysis of the Caribbean coral Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bulletin of Marine Science 55:151-175. Table 1 . Timeline of the development of the AGRRA methodology. 619 Version Date Major contributors Highlights Prototype 1 1995 Ginsburg, Bak, Kiene, Gischler and Kosmynin. (1996) Large stony corals-mortality and size quantified in Florida Keys Prototype 2 1996- 1997 Steneck and Lang Initial draft of AGRRA benthos and fish protocols prepared and field-tested in the Bahamas and Mexico Prototype Summer Kramer, First extensive field test and consistency 3 1997 Kosmynin, Marks, and Kramer training in Florida Keys and Andros; Benthos protocol-damselfish and sources of coral mortality added; Fish protocol-Roving Diver Technique used in addition to belt transects Version 1.0 Fall 1997 Steneck, Lang, Kramer & Ginsburg First version posted at AGRRA web site Version June Ginsburg, Kramer, Revised on basis of Miami June 1998 2.0 1998 Lang, Sale and Steneck Workshop: Benthos protocol-estimates of live stony coral cover added, bleaching categories defined, placement of algal quadrats and substratum type standardized, macroalgal height changed from maximum to average; Fish protocol-belt transects and Roving Diver Technique methods adopted, belt transect length and number stabilized at 30 m and 10, respectively Version March Bonaire Training Benthos protocol-minimum size for individual 2.1 1999 Workshop stony coral assessment informally changed from 25 cm to 10 cm, coral disease terminology standardized; Fish protocol-parrotfish and grunts <5 cm in total length removed; UW data sheets standardized; Excel data spreadsheets introduced Version May Akumal Training Benthos protocol-informal clarification that 2.2 1999 Workshop thin layers of sediment only should be removed from substratum in algal quadrats; Fish protocol-species in “Other” category of belt transects standardized; First Spanish translation of workshop training manual 620 Table 1 Continued. Version Date Version May 3.0 2000 Major Highlights contributors Revised on basis of Miami May 2000 Workshop: Benthos protocol-minimum number of transects set at 6, minimum size for individual stony corals changed from 25 cm to 10 cm and minimum number of colonies assessed changed from 100 to 50, minimum number of quadrats changed from 50 to 30, turf algae removed and macroalgae partitioned into fleshy and calcareous with separate measurements of their average heights, relative abundance of remaining algal functional groups redefined to be assessed without removal of canopy layers or sediment, coral recruit counts changed from optional to required, a measure of maximum reef relief added; Fish protocol-species of balistids and serranids in belt transects standardized 621 Table 2. List of stony corals surveyed in the AGRRA belt transects during 1997-2000. Suborder Family' Species' Milleporina Milleporidae Millepora alcicornis M. brazilienses^ M. complanata Scleractinia Acroporidae Acropora cervicornis A. palmata A. prolifera Agaricidae Agaricia agaricites A. fragilis A. grahamae A. humilis A. lamarcki A. tenuifolia A. undata Astrocoeniidae Caryophyllidae Faviidae Leptoseris cucullata Stephanocoenia intersepta Eusmilia fastigiata Colpophyllia natans (OR C. amaranthus, C. breviserialis and C. natans) Cladocora arbuscula Diploria clivosa D. labyrinthiformis D. strigosa Favia fragum F. leptophylla^ Manicina areolata Montastraea annularis (OR Montastraea annularis f annularis) M. cavernosa M. faveolata (OR Montastraea annularis f faveolata) M. franksi (OR Montastraea annularis f franksi) Solenastrea bournoni Meandrinidae S. hyades Dendrogyra cylindrus Dichocoenia stokesi Meandrina meandrites Mussidae Isophyllastrea rigida Isophyllia sinuosa Mussa angulosa Mussismilia braziliensis^ M. hartif M. hispida^ Mycetophyllia aliciae M. ferox M. lamarckiana (OR M. danaana and M. lamarckiana) 622 Table 2 Continued. Suborder Family' Species' Scolymia cubensis S. lacera Oculinidae Oculina sp. Pocilloporidae Madracis decactis M. formosa M. mirabilis M. pharensis Poritidae Porites astreoides P. branneri P. furcata (OR P. porites f furcata) P. porites (OR P. porites f porites) Siderastreidae Siderastrea radians S. siderea (OR Siderastrea radians f siderea) S. stellata^ ' Family and species names as in Cairns et al. (1991), except for Foster’s (1987) revision of Stephanocoenia, and Weil and Knowlton’s (1994) revision of the Montastraea annularis species complex. ^ in Brazil 623 Table 3. List of fishes surveyed in the AGRRA belt transeets (Version 2.2). Family names’ Species names’ Scientific Common Scientific Common Acanthuridae surgeonfish Acanthurus bahianus ocean surgeonfish A. chirurgm doctorfish A. coeruleus blue tang Balistidae leather] ackets A lu ter us script us scrawled filefish B. vetula queen triggerfish Cantherhines macrocerus whitespotted filefish C. pullus orangespotted filefish C. stifflamen ocean triggerfish Melichthys niger black durgon Xanthichthys ringens Sargassum Triggerfish Chaetodontidae butterflyfish Chaetodon aculeatus longsnout butterflyfish C. capistratus foureye butterflyfish Chaetodon ocellatus spotfin butterflyfish Chaetodon sedentarius reef butterflyfish Haemulidae^ Chaetodon striatus banded butterflyfish grunt Anisotremus surinamensis black margate A. virginicus porkfish Haemulon album margate H. aurolineatum tomtate H. carbonarium caesar grunt H. chrysargyreum smallmouth grunt H. flavolineatum French grunt H. macrostomum Spanish grunt H. melanurum cottonwick H. parra sailors choice H. plumieri white grunt H. sciurus bluestriped grunt H. striatum striped grunt Lutjanidae snapper Lutjanus analis mutton snapper L. apodus schoolmaster L. cyanopterus cubera snapper L. griseus gray snapper L jocu dog snapper L. mahogoni mahogany snapper L synagris lane snapper Ocyurus chrysurus yellowtail snapper Pomacanthidae angelfish Centropyge argi cherubfish Holacanthus bermudensis blue angelfish H. ciliaris queen angelfish H. tricolor rock beauty Pomacanthus arcuatus gray angelfish P. paru French angelfish 624 Table 3 Continued. Family names’ Species names’ Scientific Common Scientific Common Scaridae^ parrotfish Scarus coelestinus midnight parrotfish S.coeruleus blue parrotfish S. croicensis striped parrotfish S. guacamaia rainbow parrotfish S. taeniopterus princess parrotfish S. trispinosus^ greenlip parrotfish S. vetula queen parrotfish Sparisoma atomarium greenblotch parrotfish S. aurofrenatum redband parrotfish S.chrysopterum redtail parrotfish S. radians bucktooth parrotfish S. rubripinne redfin parrotfish S. viride stoplight parrotfish Serranidae sea basses Epinephelus adscensionis rock hind (grouper) E. cruentatus graysby E. fulvus coney E. guttatus red hind E. marginatus^ dusky grouper E. striatm Nassau grouper Mycteroperca bonaci black grouper M interstitialis yellowmouth grouper M. rubra comb grouper M. tigris tiger grouper M. venenosa yellowfm grouper Others Bodianus rufus Spanish hogfish Caranx ruber bar jack Lachnolaimus maximus hogfish Microspathodon chrysurus yellowtail damselfish Sphyraena barracuda great barracuda ’Nomenclature as in Robbins et al. (1991); see Appendix Two for the corresponding generic and specific names in Eschmeyer et al. (1998). ^ Excluding Haemulidae and Scaridae <5 cm in total length. ^ in Brazil APPENDIX TWO FISH BIOMASS CONVERSION EQUATIONS BY KENNETH W. MARKS* and KRISTI D. KLOMP^ INTRODUCTION Fish abundance determined from the AGRRA belt transects has been reported throughout this volume as density estimates (number of individuals/lOOm^). Most investigators using the AGRRA protocol have utilized a consistent methodology for censusing fish so that abundance estimates reported as density are especially useful in making regional comparisons of fish abundance. In some instances, however, where individual species within the reported taxa have widely varying morphologies, biomass (weight) may be a more representative measure of fish abundance. Biomass is an important attribute of populations that may be of interest to ecologists and resource managers since it provides insight into the trophic structure of the community and the production capacity of a reef (Bohnsack and Harper, 1988; Anderson and Neumann, 1996). Some of the authors in this volume have chosen to present their fish abundance results as biomass (grams/ lOOm^). Fish weight was estimated using previously established length- weight relationships for Caribbean fishes (Table 1). METHODS Fish weight was calculated using the power function: W = aL^, where W is the weight (grams), L is the length (cm), and a and b are parameters estimated by linear regression of logarithmically transformed length-weight data. The parameters a and b shown in Table 1 were adjusted for unit length from linear regressions performed on fish lengths (mm) reported previously in the literature. Most of the length-weight relationships were determined from southern Florida specimens (Bohnsack and Harper, 1988), with exceptions as noted from Bohnsack and Harper (1988), Bullock et al. (1992), Claro and Garcia- Arteaga (1994), and Letoumeur et al. (1998). ’ Reef Environmental Education Foundation, 98300 Overseas Highway, Key Largo, FL, 33037. Email: kema@adelphia.net Department of Fisheries and Wildlife, Michigan State University, East Lansing, MI 48824, USA. Email: klompkri@msu.edu 626 REFERENCES Anderson, R.O., and R.M. Neumann 1996. Length, weight, and associated structural indices. Pp. 447-482. In: B.R. Murphy and D.W. Willis (eds.). Fisheries Techniques. Second edition. American Fisheries Society. Bethesda, MD. Bohnsack, J.A., and D.E. Harper 1 988. Length-weight relationships of selected marine reef fishes from the southeastern United States and the Caribbean. NOAA Technical Memoir NMFS-SEFC-215, 31 pp. Bullock, L.H., M.D. Murphy, M.F. Godcharles, and M.E. Mitchell 1992. Age, growth, and reproduction of jewfish Epinephelus itjara in the eastern Gulf of Mexico. Fishery Bulletin 90:243-249. Claro, R., and J.P. Garcia-Arteaga 1994. Crecimiento. Pp. 321-402. In: R. Claro (ed.), Ecologla de los Feces Marinos de Cuba. Instituto de Oceanologia, Academia de Ciencias de Cuba and Centro de Investigaciones Quintana Roo (CIQRO) Mexico. Eschmeyer, W.N. (editor) 1998. Catalog of Fishes. Special Publication No. 1 of the Center for Biodiversity Research and Information, California Academy of Science, San Francisco, CA. Volumes 1-3, 2905 pp. Letoumeur, Y., M. Kulbicki, and P. Labrosse 1998. Length-weight relationships of fish from coral reefs and lagoons of New Caledonia, southwestern Pacific Ocean: an update. Naga, ICLARM Q. 21(4):39-46. Table 1. Length-weight relationships for the AGRRA fishes. 627 628 629 Plate 13A. Reef fishes play many important roles in coral reef community dynamics by their trophic interactions as herbivores and as predators. Schooling acanthurids, as shown here, overwhelm damselfish to raid their gardens. In the AGRRA belt transect method, sampling biases are minimized by restricting the width of the belt transect and the number of species recorded which help to maintain a relatively consistent search image. (Photo Kenneth W. Marks) Plate 13B. Belt transects are used to estimate the abundance and size (used for biomass estimations) of ecologically and/or commercially significant fishes, such as these black groupers (Mycteroperca bonaci) which have been overharvested in much of the wider Caribbean. REEF’s Roving diver surveys are used to assess fish species richness and relative abundance and complement the belt transect method. (Photo Robert W. Steneck) 630 Plate 14A. One of the main objectives of the AGRRA approach is to provide a standardized methodology enabling teams working in different areas to collect and compare data on a regional scale. The transect-based benthos protocol is focused on several indicators of the condition of stony corals and the abundance of reef algae and Diadema. (Photo Kenneth W. Marks) Plate 14B. Two distinct methods, belt transects as shown here, and REEF roving diving surveys, provide complementary “snapshots” of fishes at AGRRA assessment sites. (Photo Kenneth W. Marks) ,ippji(vfiflJ» f*1(,(ir|;ii>v|tli? w^i rrAwikjcd it) dftf<5r(vtuJ'(,i» v-aifjilli #4 (Hit>t «• rl^w *i» Ait' jthos.itlu'rri. «M(d Si rv>?t^^I>T ot'4n;f. ?\Yi!( afin^:ii»tnftWyiwi *H Ai»'3h«)s.i»1ii'rr*. f«(d Si u' 1 ■■>T hV} m- |SFf; :,-.£: D ■)